00Harrison_FM.

qxd

2/25/05

12:38 PM

Page i

PRAISE

FOR

LINUX QUICK FIX NOTEBOOK

“Where has this book been all my professional life? It is about time someone wrote this book. The author covers many basic, practical issues that other authors omit in a format that makes for a great troubleshooting reference and for casual perusing. There are a lot of books available that give a broad overview of system and network administration, and there are a lot of detailed books on each specific topic covered in this book, but I don’t know of any other place where I can get so much useful, practical information on such a wide range of topics in such a concise and readable form. I will highly recommend it to my friends and colleagues who are trying to manage their own systems. I am anxious to show it to our instructors who teach data communications and networks as a possible text or lab manual.” —Ron McCarty, instructor and program chair, Computer Science, Penn State Erie “This book takes a very unique approach and is very flexible. It offers a great deal to the home user, students, small businesses—all the way to any Linux sys admin that needs a handy reference. I found it very refreshing in terms of style, stated goals, presentation, ease of use, direction, and the “present and future” value of the content. In other words the author is very current and knowledgeable. I would recommend this book to every level from programmer to manager.” —Joe Brazeal, Information Technician III, Southwest Power Pool “This book is an excellent resource for new Linux administrators. I was impressed with the comprehensive set of 35 topic chapters, making this book a very useful reference. I know of no other work like this.” —George Vish II, Linux curriculum program manager and senior education consultant, Hewlett-Packard Education “I would recommend this book to novice Linux users and administrators. It covers a lot of topics, and makes it easy for readers to follow along.” —Bret Strong, education consultant, Hewlett-Packard Company “Linux Quick Fix Notebook is a well-written guide to designing, installing, and maintaining Linux servers. Many Linux reference texts are just simple rewrites of Linux ¨man¨ and ¨info¨ pages, but this book addresses the real issues that systems administrators encounter when working with Linux. The author provides a pedagogical guide to learning Linux server administration. In addition, the book includes step-by-step information so that users can quickly take advantage of the best that open source software has to offer today. I would recommend this book to both beginning and intermediate users of Linux.” —Joseph Naberhaus, Professor of Networking, Universidad Del Mar

00Harrison_FM.qxd

2/25/05

12:38 PM

Page ii

00Harrison_FM.qxd

2/25/05

12:38 PM

Page iii

Linux Quick Fix Notebook
®

00Harrison_FM.qxd

2/25/05

12:38 PM

Page iv

BRUCE PERENS’ OPEN SOURCE SERIES
http://www.phptr.com/perens

◆ C++ GUI Programming with Qt 3

Jasmin Blanchette, Mark Summerfield

◆ Managing Linux Systems with Webmin: System Administration and

Module Development Jamie Cameron Mel Gorman

◆ Understanding the Linux Virtual Memory Manager ◆ Implementing CIFS: The Common Internet File System

Christopher Hertel

◆ Embedded Software Development with eCos

Anthony Massa

◆ Rapid Application Development with Mozilla

Nigel McFarlane

◆ The Linux Development Platform: Configuring, Using, and Maintaining a

Complete Programming Environment Rafeeq Ur Rehman, Christopher Paul

◆ Intrusion Detection with SNORT: Advanced IDS Techniques Using SNORT,

Apache, MySQL, PHP, and ACID Rafeeq Ur Rehman

◆ The Official Samba-3 HOWTO and Reference Guide

John H. Terpstra, Jelmer R. Vernooij, Editors

◆ Samba-3 by Example: Practical Exercises to Successful Deployment

John H. Terpstra

00Harrison_FM.qxd

2/25/05

12:39 PM

Page v

Linux Quick Fix Notebook
®

Peter Harrison

Upper Saddle River, NJ • Boston • Indianapolis • San Francisco New York • Toronto • Montreal • London • Munich • Paris Madrid • Capetown • Sydney • Tokyo • Singapore • Mexico City

00Harrison_FM.qxd

3/1/05

4:08 PM

Page vi

Many of the designations used by manufacturers and sellers to distinguish their products are claimed as trademarks. Where those designations appear in this book, and the publisher was aware of a trademark claim, the designations have been printed with initial capital letters or in all capitals. The author and publisher have taken care in the preparation of this book, but make no expressed or implied warranty of any kind and assume no responsibility for errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of the use of the information or programs contained herein. The publisher offers excellent discounts on this book when ordered in quantity for bulk purchases or special sales, which may include electronic versions and/or custom covers and content particular to your business, training goals, marketing focus, and branding interests. For more information, please contact: U. S. Corporate and Government Sales (800) 382-3419 corpsales@pearsontechgroup.com For sales outside the U. S., please contact: International Sales international@pearsoned.com Visit us on the Web: www.phptr.com Library of Congress Number: 2004117224 Copyright © 2005 Pearson Education, Inc. Printed in the United States of America. This material may be distributed only subject to the terms and conditions set forth in the Open Publication License, v1.0 or later (the latest version is presently available at http://www.opencontent.org/openpub/). ISBN 0-13-186150-6 Text printed in the United States on recycled paper at R.R. Donnelley and Sons Company in Crawfordsville, Indiana. First printing, March 2005

00Harrison_FM.qxd

2/25/05

12:39 PM

Page vii

Dedicated to Diana “Turn off the light and come to bed.” Blaze and Jade “Daddy, can we stop being quiet now?” Jack and Barbara By my side every day.

00Harrison_FM.qxd

2/25/05

12:39 PM

Page viii

00Harrison_FM.qxd

2/25/05

12:39 PM

Page ix

Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xxxi Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xl About the Author . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .xli

Part 1

The Linux File Server Project

1 Why Host Your Own Site? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .3 Our Network . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Alternatives to In-House Web Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .4 Virtual Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Dedicated Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5 Factors to Consider Before Hosting Yourself . . . . . . . . . . . . . . . . . . . . . . . . . .6 Is In-House Hosting Preferred? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .6 How to Migrate From an External Provider . . . . . . . . . . . . . . . . . . . . . . . . . . .8 In-House Server Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .8 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .10 2 Introduction to Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .11 The OSI Networking Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .12 An Introduction to TCP/IP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .13 TCP Is a Connection-Oriented Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . .14 How TCP Establishes a Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .14 UDP, TCP’s “Connectionless” Cousin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 TCP and UDP Ports . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .15 The TCP/IP Time to Live Feature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16 The ICMP Protocol and Its Relationship to TCP/IP . . . . . . . . . . . . . . . . . .16

ix

00Harrison_FM.qxd

2/25/05

12:39 PM

Page x

x

Contents

How IP Addresses Are Used to Access Network Devices . . . . . . . . . . . . . . .17 Private IP Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .17 The localhost IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .18 Network Address Translation (NAT) Makes Private IPs Public . . . . . . . .18 Port Forwarding with NAT Facilitates Home-Based Web Sites . . . . . . . .19 DHCP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .19 How DNS Links Your IP Address to Your Web Domain . . . . . . . . . . . . . . .20 IP Version 6 (IPv6) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .21 How Subnet Masks Group IP Addresses into Networks . . . . . . . . . . . . . . . .21 Calculating the Number of Addresses Assigned to a Subnet . . . . . . . . . . .22 Calculating the Range of Addresses on Your Network . . . . . . . . . . . . . . . .22 Subnet Masks for the Typical Business DSL Line . . . . . . . . . . . . . . . . . . .24 The Physical and Link Layers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .24 Networking Equipment Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 Network Interface Cards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .25 The Two Broad Types of Networking Equipment . . . . . . . . . . . . . . . . . . .27 Connectivity Using Hubs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .28 Using Switches as a Faster Alternative to Hubs . . . . . . . . . . . . . . . . . . . .29 Local Area Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 How Routers Interconnect LANs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .29 How Simple Routing Works . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .30 Default Gateways, the Routers of Last Resort . . . . . . . . . . . . . . . . . . . . . .31 Firewalls Help Provide a Secure Routing Environment . . . . . . . . . . . . . .31 Additional Introductory Topics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 The File Transfer Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 Linux Help . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .32 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .33 3 Linux Networking . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .35 How to Configure Your NIC’s IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Determining Your IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .36 Changing Your IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .37 Multiple IP Addresses on a Single NIC . . . . . . . . . . . . . . . . . . . . . . . . . . .39 IP Address Assignment for a Direct DSL Connection . . . . . . . . . . . . . . . .40 IP Address Assignment for a Cable Modem Connection . . . . . . . . . . . . . .45 How to Activate/Shut Down Your NIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 How to View Your Current Routing Table . . . . . . . . . . . . . . . . . . . . . . . . . . . .45 How to Change Your Default Gateway . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .46 How to Configure Two Gateways . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Adding Routes from the Command Line . . . . . . . . . . . . . . . . . . . . . . . . . . .47 Adding Routes with /etc/sysconfig/network-scripts/ Files . . . . . . . . . . .48 How to Delete a Route . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Changing NIC Speed and Duplex . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Using mii-tool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .49 Using ethtool . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .50

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xi

Contents

xi

How to Convert Your Linux Server into a Router . . . . . . . . . . . . . . . . . . . . . .52 Configuring IP Forwarding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .52 Configuring Proxy ARP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .53 Configuring Your /etc/hosts File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .54 The loopback Interface’s localhost Entry . . . . . . . . . . . . . . . . . . . . . . . . . .55 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .56 4 Simple Network Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .57 Sources of Network Slowness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Sources of a Lack of Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Doing Basic Cable and Link Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .58 Testing Your NIC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Viewing Your Activated Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Viewing All Interfaces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .59 Testing Link Status from the Command Line . . . . . . . . . . . . . . . . . . . . . .60 Viewing NIC Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .61 How to See MAC Addresses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .62 Using ping to Test Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . .64 Using telnet to Test Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . .65 Linux telnet Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Successful Connection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .66 Connection Refused Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 telnet Timeout or Hanging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 telnet Troubleshooting Using Windows . . . . . . . . . . . . . . . . . . . . . . . . . . . . .67 Screen Goes Blank—Successful Connection . . . . . . . . . . . . . . . . . . . . . . . .68 Connect Failed Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 telnet Timeout or Hanging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 Testing Web Sites with the curl and wget Utilities . . . . . . . . . . . . . . . . . . . . .68 Using curl . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .68 Using wget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .69 The netstat Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .70 The Linux iptables Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Determining Whether iptables Is Running . . . . . . . . . . . . . . . . . . . . . . . .71 How to Stop iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 How to Configure iptables Rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Using traceroute to Test Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .71 Sample traceroute Output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 Possible traceroute Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .72 traceroute Time Exceeded False Alarms . . . . . . . . . . . . . . . . . . . . . . . . . . .73 traceroute Internet Slowness False Alarm . . . . . . . . . . . . . . . . . . . . . . . . .74 traceroute Dies at the Router Just Before the Server . . . . . . . . . . . . . . . .75 Always Get a Bidirectional traceroute . . . . . . . . . . . . . . . . . . . . . . . . . . . .75 ping and traceroute Troubleshooting Example . . . . . . . . . . . . . . . . . . . . . .77 traceroute Web Sites . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .77 Possible Reasons for a Failed traceroute . . . . . . . . . . . . . . . . . . . . . . . . . .78

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xii

xii

Contents

Using MTR to Detect Network Congestion . . . . . . . . . . . . . . . . . . . . . . . . . . .79 Viewing Packet Flows with tcpdump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79 Analyzing tcpdump files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Common Problems with tcpdump . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .82 Viewing Packet Flows with tethereal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83 Basic DNS Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 Using nslookup to Test DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .85 Using the host Command to Test DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . .86 Using nmap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .87 Determining the Source of an Attack . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .88 Who Has Used My System? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 The last Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 The who Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .89 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .90 5 Troubleshooting Linux with syslog . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .91 syslog . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 The /etc/syslog.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .92 Activating Changes to the syslog Configuration File . . . . . . . . . . . . . . . .93 How to View New Log Entries as They Happen . . . . . . . . . . . . . . . . . . . . .93 Logging syslog Messages to a Remote Linux Server . . . . . . . . . . . . . . . . .94 syslog Configuration and Cisco Network Devices . . . . . . . . . . . . . . . . . . .96 syslog and Firewalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 logrotate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 The /etc/logrotate.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .96 Sample Contents of /etc/logrotate.conf . . . . . . . . . . . . . . . . . . . . . . . . . .97 The /etc/logrotate.d Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .97 Activating logrotate . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Compressing Your Log Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .98 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .99 6 Installing RPM Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .101 Where to Get Commonly Used RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs on Your Installation CDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs Downloaded from Fedora . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .102 RPMs Downloaded from rpmfind.net . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Getting RPMs Using Web-Based FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Using the Fedora Web site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Using the rpmfind Web site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .103 Getting RPMs Using Command-Line Anonymous FTP . . . . . . . . . . . . . . . .104 Getting RPMs Using wget . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107 Automatic Updates with yum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .107 Configuring yum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .108 Creating Your Own yum Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .109

00Harrison_FM.qxd

2/25/05

12:39 PM

Page xiii

Contents

xiii

Before You Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .110 Keeping Your System Current with yum . . . . . . . . . . . . . . . . . . . . . . . . . .110 Example of a yum Package Installation . . . . . . . . . . . . . . . . . . . . . . . . . . .112 How to Install RPMs Manually . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Using Downloaded Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 Using CD-ROMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .113 How to Install Source RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .114 RPM Installation Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 Failed Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .115 Signature Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .116 How to List Installed RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 Listing Files Associated with RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .117 Listing Files for Already Installed RPMs . . . . . . . . . . . . . . . . . . . . . . . . .118 Listing Files in RPM Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .118 Listing the RPM to Which a File Belongs . . . . . . . . . . . . . . . . . . . . . . . . .119 Uninstalling RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Which RPMs Start at Boot Time? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Installing Software Using tar Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .119 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .120 7 The Linux Boot Process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 The Fedora Boot Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .121 Determining the Default Boot runlevel . . . . . . . . . . . . . . . . . . . . . . . . . . . . .123 Getting a GUI Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .124 Get a Basic Text Terminal Without Exiting the GUI . . . . . . . . . . . . . . . . . .125 Using a GUI Terminal Window . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Using Virtual Consoles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 System Shutdown and Rebooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Halt/Shut Down the System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .125 Reboot the System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .126 Entering Single-user Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .127 How to Set Programs to Run at Each runlevel . . . . . . . . . . . . . . . . . . . . . . .129 chkconfig Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .129 Using chkconfig to Improve Security . . . . . . . . . . . . . . . . . . . . . . . . . . . .130 Final Tips on chkconfig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .132 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .133 8 Configuring the DHCP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .135 Download and Install the DHCP Package . . . . . . . . . . . . . . . . . . . . . . . . . .135 The /etc/dhcpd.conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .136 How to Get DHCP Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 DHCP Servers with Multiple NICs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .138 Temporary Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 Permanent Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139

. . . . . . . . . . . . . . . . . . . . . . . .140 9 Linux Users and sudo . . . . . . . . . . . . .139 DHCP Clients Obtaining 169. . . . . . . .151 How SWAT Makes Samba Simpler . .142 How to Change Passwords . . . . . . .161 The [printers] Share Section . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .151 The Samba Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . .148 10 Windows. . . . . . . . . . . . . . . . . . . . . . . . . . .162 Home Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .150 How to Get Samba Started . . . . . . . . . . . . . . . . . . . . . . . .160 The [netlogon] and [profiles] Share Sections . . . . . . . .141 In This Chapter . . . . . . . . . . . . . . . . . . . . . . .162 Samba Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .162 How to Create a Samba PDC Administrator User . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154 Encrypting SWAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . .149 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 Example of a User Using sudo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .140 Conclusion . . . . . . . .145 Simple /etc/sudoers Examples . . . . . . .0 Addresses . . . . . . . . . . . . . . . . . . . . . . . .152 Basic SWAT Setup . . . . . . . . . . . . . . . . . . .152 Controlling SWAT . . . . . . . . . . . . . . . . . . . . . . . . . . .163 Make Your PC Clients Aware of Your Samba PDC . . . . . . . . . . . . . . . .0. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .142 How to Delete Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .154 How to Make SWAT Changes Immediate . . . . . . . . .146 Conclusion . . . . . . . . . . .163 Create Samba Trust Accounts for Each Workstation . . . . .158 The [homes] Section . . . . . . . . . . . . . . . Linux. . . . . . . . . . . . . .162 Corporate Environment . . . . . . . . . . . . . .143 How to Change the Ownership of a File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 How to Add Users . . . . . . . . . .163 How to Add Workstations to Your Samba Domain . . . . . . . . . . . . . . . . and Samba . . . . .164 . . . . . . . . . . . . . . . . . . . . . . . . . . . .158 The [global] Section . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xiv xiv Contents Configuring Linux Clients to Use DHCP .158 Creating a Starter Configuration . .139 Simple DHCP Troubleshooting .141 Adding Users .161 Shares for Specific Groups of Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .139 Configuring Windows Clients to Use DHCP . . . . . . . .149 Download and Install Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .141 Who Is the Super User? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .144 Using sudo . . . . . . . . . . . . . . . . . . . . . . . . . .143 How to Tell the Groups to Which a User Belongs . . . . . .144 Downloading and Installing the sudo Package . . . . . . . . . . . . . .

. . . . . . . . . . . . .175 Test Your Windows Client Configuration . . . . . . . . . . .182 Checking the Samba Logs . . . . . . .175 Sharing Windows Drives Using a Linux Samba Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168 How to Modify Samba Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .178 12 Samba Security and Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . .171 Adding a Printer to a Samba PDC . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .168 How to Delete Users from Your Samba Domain . . . .179 Samba and Firewall Software . . . .180 Windows-based Zone Alarm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 Mapping a Private Windows Drive Share . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .186 Basic Samba Security . . . . . . . . . . . . . . . . . . . . .174 Create the Directory and User Group . . . . . . .189 Encryption Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .176 Create a CD-ROM Drive Mount Point on Your Samba Server . . . . . . . . . . . . . . . . . . .180 Linux iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .179 Testing the smb. . . . . . . . . . . . . . . . . . . . . . . . . . .166 Mapping the Linux Users to an smbpassword . . .187 Wireless Linux-Compatible NICs . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Configure the Printer Driver on the Workstations . . . . . .177 Automating Mounting with Linux Samba Clients . . . . . . . . . . . . . . . . . . . . . . . . . .186 13 Linux Wireless Networking . . . . . . .181 Testing Samba Client/Server Network Connectivity . . . . . . . . . . . .171 Make Samba Aware of the Printer . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xv Contents xv How to Add Users to Your Samba Domain . . . . . . . . . . . . . . . . .167 Domain Groups and Samba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .181 The Windows XP Built-In Firewall .174 Configure the Share in SWAT . . . . . . . . . . . . . . .188 Common Wireless Networking Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .174 Map the Directory Using My Computer . . . . . . . . . . . . . . . . .169 11 Sharing Resources Using Samba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .166 Adding the Users in Linux . . . . . . . . .175 Windows Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .173 Creating Group Shares in Samba . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .188 Extended Service Set ID . . . . . . . . . . . . . . . . . .186 Samba Network Troubleshooting . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . .178 Conclusion . . . . . . . . . . . . . . .179 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .187 In This Chapter . .188 Wireless Access Points . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .171 Adding the Printer to Linux . . . . . . . . . . . . . . .186 Conclusion . . . .169 Conclusion . . . . . . .181 Testing Basic Client/Server Network Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . .221 Loading Kernel Modules Needed by iptables . . . . .203 Unknown Device Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220 Recovering from a Lost Script . . . . .196 Testing Linux-WLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .203 Can’t Ping Default Gateway . . . .qxd 2/25/05 12:39 PM Page xvi xvi Contents Networking with Linux Wireless Tools . . . . .219 Saving Your iptables Scripts . . . . . . . . . . . . . . . . .209 In This Chapter . . . .192 Installing the RPMs . . . . . . .215 Important iptables Command Switch Operations . . . . . . . . .197 Troubleshooting Your Wireless LAN . .191 Linux-WLAN Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227 . . . . . . .225 Advanced iptables Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .220 Fedora’s iptables Rule Generator . . . . . . . . . . . . . . . . . . . . . . . .203 A Common Problem with Linux-WLAN and Fedora Core 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .200 Check for Interrupt Conflicts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .196 Linux-WLAN Encryption for Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222 Basic Operating System Defense . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .193 Linux-WLAN Post Installation Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .206 Part 2 The Linux Web Site Project 14 Linux Firewalls Using iptables . .209 What Is iptables? . . . . . . . .211 How to Start iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .192 Identifying the Correct RPMs . . . . .199 Check the NIC Status . . . . . . . . . . . . . . . . . . . . . . . . . . .202 Kernel Log Errors . . . . . . . . . . . . . . . . . . . . . . . . . .216 Using User-Defined Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .222 Sample iptables Scripts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 Targets and Jumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .226 Allowing DNS Access to Your Firewall . . . . . . . . . . .191 Networking with Linux-WLAN . . . . . . . .189 Permanent Wireless Tools Configuration . . . . . . . . . . . . . . . .190 Wireless Tools Encryption . . . . . . . . . .211 Determining the Status of iptables . . . . .210 Download and Install the iptables Package . . . . . . . . . . . . . . . . . . . .204 Wireless Networks In Businesses . . . . . . . . . . . . . . . . . . . . . . . . . . . . .211 Packet Processing in iptables . . . . . . . . . . . . . . . . . .223 Basic iptables Initialization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .227 Allowing WWW and SSH Access to Your Firewall . . . . . . .205 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .189 Using iwconfig for Wireless Tools Configuration . . . . . . . . . . . . . . .194 Start Linux-WLAN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

.qxd 2/25/05 12:39 PM Page xvii Contents xvii Allowing Your Firewall to Access the Internet . . . . . . . . . . . . . . . . . . . .245 The /etc/vsftpd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .243 The vsftpd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .230 Static NAT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .249 16 Telnet. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .236 15 Linux FTP Server Setup . . . . . . . . . . . . . . . . . . . . . . .251 Controlling xinetd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .240 Client Protected by a Firewall Problem . . . . . . . . .234 Checking the Firewall Logs . .229 Port Forwarding Type NAT (DHCP DSL) . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245 FTP Greeting Banner . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .234 iptables Won’t Start . . . . . TFTP. . . . .247 Sample Login Session to Test Functionality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242 Testing the Status of VSFTPD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .245 FTP Security Issues . . . . . . . . . . . . . . .237 FTP Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254 Basic TELNET Security . . . . . . . .232 Troubleshooting iptables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .252 Using The TELNET Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .ftpusers File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251 In This Chapter .254 . . . . . . . . . . . . . . .236 Conclusion . . . . . . . . . . . . . . . . and xinetd . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .254 Stopping a TELNET Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .251 Controlling xinetd-Managed Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .242 How to Get VSFTPD Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf Options . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 Tutorial . . . . . . . . . . . . . . . . . . .248 Conclusion . . .237 Types of FTP . . . . . . . . . . . . . . .251 Managing xinetd Programs . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .229 Masquerading (Many to One NAT) . . . . . . . . . . . . . . . .245 Anonymous Upload . . . . . . . . . . . . . . . . . . . . . . . . . .246 Using SCP as Secure Alternative to FTP . . . . . .243 Other vsftpd. . . . . . .252 TELNET . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .238 Problems with FTP and Firewalls . . . . . . .253 Setting Up a TELNET Server . . .241 How to Download and Install VSFTPD . . . . . . . . . . . . . . .240 Server Protected by a Firewall Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 Troubleshooting FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .228 Allow Your Home Network to Access the Firewall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .246 FTP Users with Read-Only Access to a Shared Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .252 Installing the TELNET Server Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . .237 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276 DNS Clients . . . . . . . . . . . . . . .266 Using SSH to Log Into a Remote Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .275 In This Chapter . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . .263 Starting OpenSSH . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .272 Configuration: Server Side . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .265 SSH Versions 1 and 2 . . . . . . . . . . . . . . . . . . . . . . . . .265 The /etc/ssh/sshd_config File . . . . . . . . . . . . . . . . . . . . .257 Saving Cisco Configurations to the TFTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279 . . . . . . . . .263 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .269 Copying Files to the Local Linux Box . . . . . . . . . . . . . . . . . . . . .273 Conclusion . .275 Introduction to DNS . . .268 Executing Remote Commands on Demand with SSH .276 Authoritative DNS Servers . . . .276 How DNS Servers Find Your Site Information . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .262 17 Secure Remote Logins and File Copying . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xviii xviii Contents TFTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .277 When To Use A Dynamic DNS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . .258 Uploading Cisco Configurations from the TFTP Server . . .276 BIND . . . . . . . . . . . . . . . . . . . .277 Basic DNS Testing of DNS Resolution . . . . . . .261 Conclusion . . . . . . . . . .271 Configuration: Client Side . . . . . . . . . . . . . . . . . . .277 How to Get Your Own Domain . . . . . . . . . . .260 Using TFTP to Restore Your Router Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .271 Using SSH and SCP Without a Password . . . . . . . . . . . . . . . . . . . . .273 18 Configuring DNS . . .266 Change the TCP Port on Which SSH Listens . . . . . . .275 DNS Domains . . . . .276 When to Use a Static DNS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279 Starting BIND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .263 A Quick Introduction to SSH Encryption . . . . .270 Copying Files to the Remote Linux Box . . . . . .270 SFTP: Another Secure Alternative to FTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .276 When to Use a DNS Caching Name Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .256 Installing the TFTP Server Software . . . . . . . . . . . . . .265 Testing the Status of SSH . . . . . . . . . . . . . . . . . . . . . .267 What to Expect with Your First Login .257 Configuring the TFTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .268 Deactivating TELNET After Installing SSH . . . . . . . . . . . . . . . . . . . . . . . . . . .267 SSH Failures Due to Linux Reinstallations . . . . .269 SCP: A Secure Alternative to FTP . . . . . . . . . . . . . . . . .278 Downloading and Installing the BIND Packages . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

294 General Troubleshooting Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .310 Conclusion . . . . . . . . .303 Dynamic DNS Preparation . . . . . . . . . . . . . . . . . . . . . . .307 The /etc/ddclient. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 Configuring the Zone Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .284 Sample Forward Zone File . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xix Contents xix The /etc/resolv. . . . . . . . . .307 DDNS Client Software: Linux DDclient . . . . .308 How to Get DDclient Started . . . . . . . . . . . . . . . . . . . .306 Dynamic DNS and NAT Router/Firewalls . . . . . . . . . . . . . . . . . .313 Starting Apache . . . .307 DDNS Client Software: SOHO Router/Firewalls . . . . . . . . . . . . . . . . . .conf File . . . .310 Testing Your Dynamic DNS . . . . . .292 Make Sure Your /etc/hosts File is Correctly Updated . . . . . . . .301 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .300 Naming Convention Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .313 Downloading and Installing the Apache Package . . . . . . . . . . . . . . . . . . . . . . . . . . .305 Install a DDNS Client on Your Server . . . . . . . . . . . . . . . . . .313 In This Chapter . .314 . . . . . . . . . . . . . . . . . . . . . . . . .314 Configuring DNS for Apache . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .289 Loading Your New Configuration Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .295 Migrating Your Web Site In-House . . . . . . . . . . . . . . . . . . . . .314 DHCP and Apache . . . . . . . . . . . . . . . . . . . . . . . .292 Configure Your Firewall . . . . . . . . . . . . . . . . .304 Registering DDNS . . . . . .293 Troubleshooting BIND . . . . . . . . . . .293 Fix Your Domain Registration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .288 What You Need to Know About NAT and DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .303 In This Chapter . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . .conf . . . . . . . . . . . . . . . . . .304 Testing ISP Connectivity for Your Web Site . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .282 Configuring named. . . . . . . .287 Sample Reverse Zone File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . .281 Configuring a Regular Name Server . . . . . . .309 Finding DDclient Help . . .282 Configuring resolv. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .300 Zone Transfer Protection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .280 Important File Locations . . . . . . . . . . . . . . .298 DHCP Considerations for DNS . . . . . . . . . . . . . . . . . . . . . . . . . . . .311 20 The Apache Web Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .279 Configuring a Caching Name Server . . . . . . . . . . . . . . .301 19 Dynamic DNS . . . . . . . . . . . . . . . . . . . . . .299 Simple DNS Security . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf Files When Upgrading . . . . . . . . . . . .337 How sendmail Works . . . . . . .323 Testing Your Web Site Before DNS Is Fixed . . . . . . . . . . . . . . . . . . . . .340 Starting sendmail . . . . . . . . . .329 Protecting Web Page Directories with Passwords . . . . .329 Step 1: Configure Virtual Hosting on Multiple IPs . . . . . . . . . . . . . . . . .345 The /etc/mail/access File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .326 Handling Missing Pages . . .330 The /etc/httpd/conf. . . . . . . . . . . . . . . . . . . . . . .322 Configuration: Multiple Sites and IP Addresses . . . . . . . . . . . . . . . . . . . .347 Which User Should Really Receive the Mail? . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337 Configuring sendmail . . . . . . . . . .342 How to Configure Linux sendmail Clients . . . . . . . . . . . . .332 Testing Basic HTTP Connectivity . . . .335 21 Configuring Linux Mail Servers . . . . . . . .327 Compression Configuration Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .320 IP-Based Virtual Hosting . . . . . . . . . . . . . .327 Using Data Compression on Web Pages . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . .343 Converting From a Mail Client to a Mail Server .341 Configuring DNS for sendmail . . . . . . . . . . . . . . . . . . . . .350 Using sendmail to Change the Sender’s E-mail Address . . . . . . . . . . . . . .325 Disabling Directory Listings . . . . . . . . . . . . .332 Only the Default Apache Page Appears . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .328 Apache Running on a Server Behind a NAT Firewall . . . . . . . . . . . .337 Installing sendmail . . . . . . . . . . . . . . . . . .332 Browser 403 Forbidden Messages . . . . . . . . . . . . . . . . . . . . .346 The /etc/mail/local-host-names File . . . . . . . . . . . . . . . . .321 Virtual Hosting and SSL . . . . . . . .353 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .331 Troubleshooting Apache . . . . . . . . . . . . . . . . . . . . . . . . . . .315 Where to Put Your Web Pages . . . . . . . . .335 Conclusion . . . .315 Named Virtual Hosting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .332 Incompatible /etc/httpd/conf/http. . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xx xx Contents General Configuration Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .340 The /etc/mail/sendmail. . . . . . . . . . . . . . . . . .333 The Apache Error Log Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .347 sendmail Masquerading Explained . . . .329 Step 2: Configure DNS Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .mc File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .333 Server Name Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .d Directory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .343 The /etc/mail/relay-domains File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .352 Troubleshooting sendmail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .337 In This Chapter . . . . . . . . . . . . . . . . . . .333 The Apache Status Log Files . . . . . . . . . . .340 How to Restart sendmail After Editing Your Configuration Files . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .368 Doing SNMP Queries . . . . . . . . . . . . . . . . . . . . . . .381 The Webalizer Configuration File . . .368 SNMP Versions . . . . . . . . . . . . . .385 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .371 Simple SNMP Troubleshooting . . . . . . . . . . . . . .385 Locating and Viewing the Contents of Linux MIBs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .382 Conclusion . . . . . . . . . . . . . . . . . . . .385 Testing Your MIB Value . . . .365 SNMP Community Strings . . . . . . . . . .364 22 Monitoring Server Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .363 Conclusion . . . . . . . . . . . . . . . . . . . . .363 How to Handle Overlapping E-mail Addresses . .378 Troubleshooting MRTG . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .388 The TCP/IP Monitoring MIB . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .373 MRTG . . . . . .qxd 2/25/05 12:39 PM Page xxi Contents xxi Fighting Spam . . . . . . . . . . . . . . . . . . . . . . . .363 Configuring Secure POP Mail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .381 The vmstat Command . . . . .362 Installing Your POP Mail Server . . . .361 Configuring Your POP Mail Server . . . . . . . .363 Troubleshooting POP Mail . . . .387 Differences in MIB and MRTG Terminology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .374 Configuring MRTG . . . . . . . . .388 Manually Configuring Your MRTG File . . . . . . . . . . . . . . . . . . . . .369 Basic SNMP Security . . . . . . . . . . . . . . . . . . . . . . .362 How to Configure Your Windows Mail Programs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .365 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .359 A Simple PERL Script to Help Stop Spam . . . . . . . . . .377 How to View the MRTG Graphs in Your Web Browser . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .389 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .387 The CPU and the Memory Monitoring MIB . . . . . . . . . . . . . . . . . .373 MRTG Download and Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .383 23 Advanced MRTG for Linux . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .362 Starting Your POP Mail Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .380 How to View Your Webalizer Statistics . . . . . . . . . . . . . . . . . . . . . . .375 Getting MRTG to Poll Multiple Devices . .377 Configuring Apache to Work with MRTG . . . . . . . . . . . . . .358 Using Public Spam Blacklists with sendmail . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .379 Webalizer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .365 OIDs and MIBs . . . . . . . . .365 SNMP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .358 spamassassin . . . . .381 The top Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .382 The free Utility . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .378 Using MRTG to Monitor Other Subsystems . . . . . . . . . . . . . . . .

. . . . .407 Firewalls and NTP . . . . . . . .400 24 The NTP Server . . . . . . . .406 Cisco IOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .408 Part 3 Advanced Topics 25 Network-Based Linux Installation . . . . . . . . . .398 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . .407 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Find the OIDs . . .399 Testing The OIDs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406 Configuring Cisco Devices to Use an NTP Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . .399 Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391 Defining Global Variables . . . . . .395 Testing the Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Configuring CRON to Use the New MRTG File . . . . . . . .401 Download and Install the NTP Package . . . . . . . . . . . . . . . . . . . .403 Testing and Troubleshooting NTP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .398 Monitoring Non-Linux MIB Values . . . . . . . . . . . . .397 Creating a New MRTG Index Page to Include this File . . .qxd 2/25/05 12:39 PM Page xxii xxii Contents Parameter Formats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .395 Implementing Advanced Server Monitoring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391 Defining The MIB Target Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .406 CAT OS . . . . . . . . . . . . . . . . . .390 Options Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .411 Setting Up the Installation Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .404 Your Linux NTP Clients Cannot Synchronize Properly . . . . . . . . . . . . . . . . .405 Fedora Core 2 File Permissions . . . . . . . . . . . . . . . . .395 A Complete Sample Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .391 Scaling Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403 Doing an Initial Synchronization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403 Determining If NTP Is Synchronized Properly . . . . . . . . . . . . . . . . . . . .402 How to Get NTP Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .412 Set Up Your Web Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .402 The /etc/ntp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .414 Create a Special FTP User . . . . . .389 Legend Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . .401 In This Chapter .390 Title Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .403 Verifying NTP is Running . . . . . . . . . . . . . .400 Conclusion . . . . . . . . . . . . . . . . . . . . . . .412 Basic Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .407 Configuring a Windows NTP Client . . .414 Set Up Your FTP Server . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . .427 RAID Partitioning . . . . . . . . . . . . . . . . . . . . . .432 Edit the /etc/fstab File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .422 RAID 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435 Determining the Disk Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .424 Serial ATA Drives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .436 . . . .418 How to Run a Kickstart Installation . . . . . . . . . . . . . . .421 In This Chapter . . . . . . . . .420 26 Linux Software RAID .433 Check the Status of the New RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432 Create a Mount Point for the RAID Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . .434 27 Expanding Disk Capacity . . . . . . . . . . . . . .421 RAID Types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .421 RAID 0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .415 Configure Your DHCP Server . . . . . . . . . . . . . . . .416 The Network Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Configuring Software RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Configure RAID in Single-User Mode . . . . . . . . . . . . . . . . . . . . . . . . . .417 How to Create New Kickstart Configuration Files . . . . . . . . . . .435 In This Chapter .433 Mount the New RAID Set . . . . . . .00Harrison_FM. . . . . . . . . . . . . . .423 Before You Start . . . . . . . . .434 Conclusion . . . . . . . . . . . . . . . .425 SCSI Drives . . . . .427 Edit the RAID Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .431 Format the New RAID Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .432 Start the New RAID Set’s Driver . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .424 IDE Drives . . . . . . . . . . . . . . . . . . . . . . . . . . . .431 Confirm RAID Is Correctly Initialized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 RAID 5 . .419 Conclusion . . . . . . .430 Create the RAID Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .416 Creating a Boot CD . . . . . . . . . . . . . . . . . . . . . .431 Load the RAID Driver for the New RAID Set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxiii Contents xxiii Set Up Your NFS Server . . . .419 Booting With Your Kickstart Files . . . . . . . . . . . . . . . . . . . . . . . . . .417 Differences Between Fedora and Red Hat Installation . . . . . . . . . . . . . . . . . . . .421 Linear Mode RAID . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .426 Back up Your System First . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .435 Adding Disks to Linux .435 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .423 RAID 4 . . . . . . . . .425 Should I Use Software RAID Partitions or Entire Disks? . . . . . . . . . .417 Automating Installation with Kickstart . . . . . . . . . . . . . . . . . . . . .416 Troubleshooting the Network Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . .453 Edit the User’s Quota Information . . . . . . . . . . . . . . . . . . . . .439 Migrating Data to Your New Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .452 Get Out of Single-User Mode . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .449 28 Managing Disk Usage with Quotas . . . . . . . . . . . . . . . . . . . . . .452 Create the Partition Quota Configuration Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .462 The /etc/exports File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .462 Starting NFS on the Server . . . . . . . . . . . . . . . .454 Editing Group Quotas .455 29 Remote Disk Access With NFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .462 Configuring NFS on the Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .437 Verifying Your New Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .457 In This Chapter . . . . . . . . . . . . . . .454 Editing Grace Periods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .460 Installing nfs . . . . . . . . . . . . . . . . . . . . . . . . .458 Key NFS Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .453 Testing . . . . . . . . . . .465 Activating Modifications to the /etc/exports File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .451 Enter Single-User Mode . . . . . . . . . . . . . . . . . . . .464 Making NFS Mounting Permanent . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .463 Configuring NFS on the Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . .461 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxiv xxiv Contents Preparing Partitions on New Disks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .467 Adding a Shared Directory To An Existing Exports File . . . . . . . . . . . . . . . . . . .458 Important NFS Daemons . . .454 Other Quota Topics . . . . . .457 General NFS Rules . .443 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .451 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .455 Getting Quota Reports .463 Starting NFS on the Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .439 Putting a Directory Structure on Your New Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . or Modifying a Share . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .440 Expanding Partitions with LVM . . . . . . . . . . . . . . . . . . . . . . . . . . . . .468 . . . . . .457 NFS Operation Overview . . . . .455 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .453 Initialize the Quota Table . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .451 Setting Up Quotas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .467 Deleting. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Moving. . . . . . . . . .464 NFS and DNS . . . . . . . . . . . . . . . . .452 Remount the Filesystem . . . . . . .442 Configuring LVM Devices . . . .442 LVM Terms . . . . . . . . . . . . . . . . . . .451 Edit Your /etc/fstab File . . . . . . . .467 New Exports File . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .480 Configuring the NFS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .473 The df Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .490 NIS Slave Servers . . . .485 Start the ypbind and ypxfrd Daemons . . . . . . . . .491 Configuring NIS Clients with Slaves . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 Limiting root Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 File Permissions . . . .486 Make Sure the Daemons Are Running . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488 Start the NIS Client-Related Daemons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .475 File Locking . . . . . . . . . . . .496 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .477 30 Centralized Logins Using NIS . . . . . . .479 Scenario . .483 Edit Your /etc/yp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .481 Configuring the NIS Server . . . . . . . . . . . . . .480 Configuring the NFS Client . . . . . . . . . . . . . . . . . . . . .487 Configuring the NIS Client . . . . . . . . . . . . . . . . . . . .475 NFS Hanging . . . . . . . . . . . . . .476 Restricting Access to the NFS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . .468 automounter Map Files . . . . . . . . . . . . . . . .496 Possible Password Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .486 Adding New NIS Users . .473 Other NFS Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .474 Security . . . .495 Changing Your NIS Passwords . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .472 Troubleshooting NFS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488 Verify Name Resolution . . . . . . . . . . . . . . . . . . . . . . . . . . . .472 The showmount Command . . . . . . . . . . . . . . .479 In This Chapter . . . . . . . . . . .476 Nesting Exports . . . . . . . . . . . . . . . . . . . . . . . . . .468 The Structure of Direct and Indirect Map Files . . . . .489 Test Logins via the NIS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .473 The nfsstat Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .488 Run authconfig . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .495 When There Is a NIS Master and Slave Pair . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .489 Test NIS Access to the NIS Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .484 Initialize Your NIS Domain . . . . . . . . . . . . . . . . . . .491 Configuring NIS Slave Servers . . . . .qxd 2/25/05 12:39 PM Page xxv Contents xxv The NFS automounter .495 When There Is Only an NIS Master . . . . . . . . . . .483 Start the Key NIS Server-Related Daemons . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .471 automounter Examples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .483 Edit Your /etc/sysconfig/network File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .476 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .471 Starting automounter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .469 Wildcards in Map Files . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . .512 Test Basic Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . .509 Troubleshooting LDAP Logins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .498 31 Centralized Logins Using LDAP and RADIUS . . . . . . . . . . . . . . . .505 Configuring the LDAP Client . . . . . .502 Start the ldap Daemon . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 Common LDAP Administrative Tasks .516 Deleting LDAP Users . . . . . .501 Configuring the LDAP Server . . . . . . . .conf Configuration File . . . . . . . . . . . . . . . . . . . .517 . . . . .513 LDAP Works But Is Not Using LDAPS . . . . . . . . . . . . . . . . . . . .505 Test the LDAP Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .497 NIS Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 LDAP Users Changing Their Own Passwords . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .505 Edit the ldap. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .501 Required LDAP Server RPMs . . . . . . .506 Edit the /etc/nsswitch File . . . . . . . . . . . . . . .511 Test Using ldapsearch . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .497 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .500 Downloading and Installing the LDAP Packages . . . . . . . . . . . . . . . . .501 Create an LDAP Root Password . . . . . . . . . . . . .500 Scenario . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .507 Configuring Encrypted LDAP Communication . . . . . . . . . . . . . . . . . . . .506 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .00Harrison_FM. . . . . . . . . . . . . . . .513 Possible stunnel Errors in Fedora Core 2 . . . . . . . . . . . . . . . . . . . .502 Convert the /etc/passwd File to LDIF Format . . . . . . . . . . . . . . . . . . . . . . . . . . .506 Create Home Directories on the LDAP Client . . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxvi xxvi Contents Considerations for a Non-NFS Environment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .511 Use SSH or the Linux Console . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 Modifying LDAP Users by User root . . . . . . . . . . . . . . . . . . . . . .501 Create a Database Directory . . .515 Adding New LDAP Users . . . . . . . . . .499 In This Chapter . . . . . . . . . . . . . .502 Edit the slapd. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 LDAP bind Errors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .504 Import the LDIF Files into the Database . .512 Use the tcpdump Command . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .513 stunnel Doesn’t Appear to Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .508 Configuring the stunnel LDAP Server . . . . . . . . . . . . . . . .499 The LDAP Directory Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .514 Starting and Stopping LDAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . .503 Modify the LDIF Files . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .517 LDAP Web Management Tools . . . . . . . . . . . . . . . . . . . .501 Required LDAP Client RPMs .508 Configuring the stunnel LDAP Client . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . .534 Download and Install the Kernel Sources Package . . . . . . . . . . .533 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524 Starting Squid . . . . . . . . . .530 Troubleshooting Squid . . . . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . .542 Install the Kernel Modules . . . . . . . . .528 Manually Configuring Web Browsers to Use Your Squid Server . . . . . . . . . . . . .525 Restricting Web Access by IP Address . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .522 32 Controlling Web Access with Squid . . . . . . . . . . . . . . . . . . . . . .533 The Pros and Cons of Kernel Upgrades . . . . . . . . . . . . . . .528 Making Your Squid Server Transparent to Users . .531 33 Modifying the Linux Kernel to Improve Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524 Access Control Lists . .537 Creating a Custom Kernel . . . .543 Copy the New Kernel to the /boot Partition . . . . . . . . . . . . . . . . . . . . . . . . . . . . .520 Conclusion . . . . . . . . . . .540 Configure Dependencies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .config File . . . . . . . . . . . . . . . . . .523 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .535 How Kernel Modules Load When Booting . . . . . . . . . . . . . . . . . . . . . . . . . . . . .536 Loading Kernel Modules on Demand . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .526 Password Authentication Using NCSA . . . . . . . . . . .519 Configuring the /etc/raddb/clients. . . . . . . . . . . . . . . . . . . . . . . . .534 Kernel Modules . . . . . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .542 Compile a New Kernel . . . . . . . . . . . . . . . . . . .525 Restricting Web Access by Time . . . . . .00Harrison_FM.530 Conclusion . . . . . . . . . . . . . . . . . . .542 Edit the Makefile to Give the Kernel a Unique Name . . . . . . . . . . . . . . . . . .conf File . .qxd 2/25/05 12:39 PM Page xxvii Contents xxvii Configuring RADIUS for LDAP . . . . . . . . . . . . . . . . . . . . . . . . . . . . .529 Squid Disk Usage . . . . . . . . . . . . .539 Customizing the . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .config File .518 Starting and Stopping FreeRADIUS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .518 Configuring the /etc/raddb/radiusd. . . . . . . . . . .539 Make Sure Your Source Files Are in Order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524 The /etc/squid/squid. . .518 Configuring the /etc/raddb/users File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .539 The .526 Forcing Users to Use Your Squid Server . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .524 The Visible Host Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .542 Build the Kernel’s Modules . . . . . . . . . . . . . . . .543 . . . . . . . . . . . . . . . . . . .517 Download and Install the FreeRADIUS Packages .535 Reasons for Kernel Modules . . . . . . . . . . . . . . . . . . . . . . . . .523 Download and Install the Squid Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .520 Troubleshooting and Testing RADIUS . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .549 The Location of MySQL Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .554 MySQL Database Backup .554 Configuring Your Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549 The /etc/my. . .559 In This Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 A Common Fedora Core 1 MySQL Startup Error . . . . . . . . . . . . . . . . . . . . . . . .565 Restart Openswan . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549 Creating a MySQL Root Account . . . . . . . . . . . . . . . . . . . . . . . . . . . . .549 Accessing the MySQL Command Line . . . . . . . . . . . . . . . . . .557 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . .566 . . . . . . . . . . . . . . . . . . . . .552 Listing the Data Tables in Your MySQL Database . . . . . . . . . . . . . . . . . . . . . . . . . . . . .548 Starting MySQL . . . . . . .conf Configuration File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .561 Get the Status of the Openswan Installation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .547 In This Chapter . . . . . .557 Connectivity Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .552 Viewing Your New MySQL Databases . . . . . . . . . . . . . . .562 VPN Configuration Steps Using RSA Keys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .556 Basic MySQL Troubleshooting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544 How to Create a Boot CD . . . . . . . . . . . . . . . . . . . . . .545 Conclusion . . . . . . . .559 VPN Guidelines . . . . . . . . . . . . . . . . . . . . . . . . .560 Download and Install the Openswan Package . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .563 The /etc/ipsec. . . . . . . . . . . . . . . . . . .550 Granting Privileges to Users . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .557 Test Database Access . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .556 Very Basic MySQL Network Security . . . . . . . . . . . . . .550 Creating and Deleting MySQL Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .564 Edit the /etc/ipsec. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .560 Scenario . . . . . . . . . .547 Preparing MySQL for Applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .553 Viewing Your MySQL Database’s Table Structure . . . . . . . . . . . . . . . . . . . . . .cnf File . . . . . . . . . . . . . . . . . . . . . . . . . . .558 35 Configuring Linux VPNs . . . . . . . . . . . . . . . .555 MySQL Database Restoration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .546 34 Basic MySQL Configuration . . . . . . . . . . .562 How to Fix Common Status Errors . . . . . . . .554 Recovering and Changing Your MySQL Root Password .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .563 Obtaining RSA Keys . . . . . . .qxd 2/25/05 12:39 PM Page xxviii xxviii Contents Updating GRUB . . . . . . . . . . . . . . . . . . . . . . . . . . . . .544 Updating the Kernel Using RPMs . .544 Kernel Crash Recovery . . . . . . . . . . . . . . .553 Viewing the Contents of a Table . . . . .561 How to Get Openswan Started . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .conf File . .548 Installing MySQL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .551 Running MySQL Scripts to Create Data Tables . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . .576 Determining the TCP Daemon’s Name . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 iptables Scripts . . . . . . . . . . . . . . . . .571 Using tcpdump . . . . . . . . . .598 FTP Client Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .577 Running Linux Without a Monitor . . . . . . . . . . . . . . . . . . . .567 Possible Changes to IP Tables NAT/Masquerade Rules . . . . . . . . . . . . .572 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .573 Appendix I Miscellaneous Linux Topics . . . . . . . . . . .600 Home/Small Office Protection Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .572 Invalid Key Messages . . . . . . . . . . . . . . . . . . . . . . . . . . . . .576 The TCP Wrappers File Format . . . .570 Check the Routes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .575 Fedora Core 3 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .579 Preparing to Go Headless . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .566 Testing the New Tunnel .579 Configuration Steps . . . . . . . .592 The mail-filter. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . and Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 The mail-filter. . . . . . . . . . . . . . . . . . . .reject File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .569 Testing VPN Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scripts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .567 How to Ensure Openswan Starts When Rebooting . . . . . . . . . . . . . . . . . . . . . . . . . .583 Authentication and Encryption Methods . . . .568 Using Pre-Shared Keys (PSK) . . . . .588 Appendix II Codes. . . . . . . . . .584 Internet Key Exchange (IKE) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .593 The mail-filter Script . . . . . . . .587 ICMP Codes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .575 Linux Security with TCP Wrappers . . . . . . . . . . . . . . . . . . . . . . . . . . .599 NTP Server Script .591 Apache File Permissions Script . . . . . . . . . . . . . . . . . . . . . . . . . . .585 VPN Security and Firewalls . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .591 sendmail Spam Filter Script . . . . . . . . . . . . . . . .584 IKE’s Role in Creating Security Associations . . . . .00Harrison_FM. . . . . . . . . . . . . . . . . . . . . . . . .598 FTP Server Script . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .581 Configuration Steps .581 VPN Terms and Methods . . . . . . . . . . . . . . . . . . . . . . . .571 Check syslog Error Messages . . . . . . . . . . . . . . . . . .569 Determine the Tunnel Status . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .586 TCP/IP Packet Format . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .568 Troubleshooting Openswan . . . . . . . .580 Make Your Linux Box Emulate a VT100 Dumb Terminal . . . . . .577 Adjusting Kernel Parameters . . . . . . . . . . . . . . . . . . . . . . . .586 VPN User Authentication Methods for Temporary Connections . . . . . . . . . . .qxd 2/25/05 12:39 PM Page xxix Contents xxix Initialize the New Tunnel . . . .600 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .accept File . . . . . . . . . . . . . . . . .

. . . . . . . . . . . .623 Cisco Local Director . . . . .622 Webalizer Differences . . . . . . . . . . . . . . . . . . . . .617 FTP Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .622 Appendix IV syslog Configuration and Cisco Devices . . . . . . . . . .mc File . . .613 Sample /etc/mail/access File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .615 Appendix III Fedora Version Differences . . .conf File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .619 Wireless Networking Differences . . .613 Sample /etc/mail/local-host-names File . . . . . . . . . . . . . . . .org (External View) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .qxd 2/25/05 2:31 PM Page xxx xxx Contents Sample DNS Zone Files: Using BIND Views .617 DHCP Differences . . . . . . . . . . . . . . . . . .619 Software Installation Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .619 iptables Differences . . . . . . . . . . . . .613 Sample /etc/mail/sendmail. . . . . . . . . . . . . . . . . . . .618 Routing Differences . . . . . . .625 Index . . . .0 RPMs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .611 Reverse Zone File for a Home Network Using NAT . . . . .00Harrison_FM. . . . . . . . . . . . .611 Zone File for my-web-site.621 File Locations .621 Precedence Bitwise Error with Red Hat 9 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .624 Cisco CSS11000 (Arrowpoints) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .623 Catalyst CAT Switches Running CATOS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .620 Linux-WLAN File Locations Using Red Hat 8. . . . . . . . . . . . . . . . . .0 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .623 Cisco Routers . . . . .conf File . . . . . . . . . . . . . . . . . . . . . . .org (Internal View) . . . . . .618 DNS Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .621 indexmaker MRTG_LIB Errors with Red Hat 9 and 8. . . . . . . . . . . . . . . . . . . .613 Sample /etc/mail/virtusertable File . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .625 The Sample Cisco syslog. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .609 The /etc/named. . . .612 sendmail Samples . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .624 Cisco PIX Firewalls . . . . . . . . . . .627 . . . . . . . . . . . . . . . .620 MRTG Differences . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .609 Zone File for my-web-site. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

It is transforming information technology in many exciting ways. The encyclopedic guides for beginners focus on Linux fundamentals and only then introduce you to more specialized topics. each with its own GUI interface. To avoid confusion between the many flavors of Linux. Linux has a variety of printed and electronic guides to show you what to do.00Harrison_FM. there are few practical texts in between that help you to make the transition from being a beginner to having the confidence of an expert. The notebook also includes many of the most commonly encountered errors with explanations of their causes and how to fix them. this guide exclusively uses the command line to illustrate the tasks needed to be done. Unfortunately the “advanced sections” in these guides cover the underlying theory reasonably well. It provides all the expected screen output when configuring the most commonly used Linux applications to help assure readers that they are doing the right thing. Unfortunately.qxd 2/25/05 12:39 PM Page xxxi Preface Without question. Like its many uses. GUI interfaces. and an advanced section covering the topics most nondesktop-support IT professionals use on a day-to-day basis. ® . and text editors. from being used in products ranging from cell phones and PDAs to cars and mainframe computers. an intermediate section covering Microsoft Office clone productivity suites. Linux Quick Fix Notebook takes these “advanced sections” and expands them sufficiently to provide a practical tutorial guide on how to do basic configuration of many popular Linux back-office applications with command-bycommand instructions. The specialist guides are highly detailed. but often are short on space to adequately cover detailed configuration instructions. WHY IS THIS BOOK NECESSARY? Most Linux “encyclopedias” are split in three sections: an introductory section covering topics such as CD-based Linux installation. focusing on narrow areas of excellence. Linux is rapidly becoming the operating system of choice in many core areas of business. IT professionals frequently have to purchase additional specialist books on each topic.

basic Windows exposure to the concept of sharing directories between servers is needed. many of the introductory topics are not covered. the third section covers more advanced topics that will become invaluable as your Linux administration role expands Explanations are given not as a lecturer. manage. and IT professionals who often have to do advanced tasks in which the underlying theory is understood. The book creates a typical departmental. power users. and every chapter has real-world practical examples in the form of tutorials.qxd 2/25/05 12:39 PM Page xxxii xxxii Preface PREREQUISITES The book’s format is aimed at proficient beginners. community college and trade school professors and students. and monitor your own Linux-based Web site running on a simple DSL or cable modem line. The first starts as an introduction to networking and extends into using Linux as a main departmental server. The chapters have a logical flow of information starting with concise backgrounders and ending with a troubleshooting section. computer literate business owners. A great deal of attention has been paid to troubleshooting techniques that are often needed to remedy unexpected behavior. or home network and shows you how to set up the Linux servers most businesses need. but the commands to do it are forgotten or at the tips of their tongues. Three sections have been created to make this process easier. students. but as a trusted and experienced coworker. To maintain its appeal as a compact guide. APPROACH With this in mind. small-office.00Harrison_FM. which provides room for much more coverage of the steps needed to get the more difficult jobs done. Web. The next section expands on this knowledge to show you how to create. Additionally. Finally. Because the readers are assumed to be exposed to the theory of Linux. the persons most likely to be interested in this book would be IT professionals and consultants. and DNS server setup Web sites on DHCP Internet links Time synchronization with NTP Error reporting with syslog Restricting users’ disk space usage with quotas . only the essential supporting theory is provided to help end users implement their projects under budget and ahead of schedule. They cover these essential topics: Software installation Networking setup and troubleshooting Samba for Windows files on Linux servers Linux-to-Linux file sharing with NFS Simple MySQL database administration LDAP and NIS for centralized logins FTP and SCP file transfers Disk drive redundancy with software RAID Wireless Linux networks MRTG server performance monitoring Linux firewalls and VPNs Squid for Web access control Mail. and SOHO workers. Two to three months of hands-on Linux experience is an ideal prerequisite.

. In summary.qxd 2/25/05 12:39 PM Page xxxiii Preface xxxiii Many of the topics are covered in Linux certification exams making the book a valuable study guide for those seeking new areas of professional development. This is the guide that will give you enough time to eat lunch. and as Linux and Windows gain equal footing in business. as computers move from the data center to the desktop. it becomes harder to remember and do it all. this book ☞ Shows how to expand inexpensively an existing IT investment in Windows using Linux as the anchor of a network and Web site ☞ Shows how to do this command by command ☞ Is highly focused on being task-oriented ☞ Illustrates how to create a simple network for small business.00Harrison_FM. corporate departments. and homes ☞ Provides an excellent networking familiarization and troubleshooting guide As the line between power users and administrators continues to blur.

Chapter 2 Introduction to Networking Covers the basic concepts needed to make the advanced user comfortable with many frequently encountered networking scenarios. Chapter 7 The Linux Boot Process Explanations of how Linux starts up after turning on the power switch. Chapter 4 Simple Network Troubleshooting Very detailed yet simply explained coverage of the various methods you can use to identify network trouble spots. Coverage of how to activate/deactivate the startup of various important system programs when the system boots up.qxd 2/25/05 12:39 PM Page xxxiv xxxiv Preface BREAKDOWN PART 1 OF BOOK The Linux File Server Project Chapter 1 Why Host Your Own Site? A discussion on the pros and cons of hosting your own Web site. Chapter 6 Installing RPM Software All you need to know on installing Linux software using popularly available RPM installation files. Chapter 3 Linux Networking A practical guide to configuring a network interface and simple Internet routing using Linux.00Harrison_FM. . Targets usage in the simple networks found in small offices and schools. Describes 30 of the most commonly used terms with simple one-paragraph discussions. Also includes a discussion on how to simply migrate a Web site from a third-party provider to your office or home. Details on how to choose between using the graphical and regular text interfaces on the Linux VGA console. Chapter 5 Troubleshooting Linux with syslog Explanations on how to view and use Linux system error messages as part of the troubleshooting process. Chapter 8 Configuring the DHCP Server Essential topics on how to configure a DHCP server so that PCs are automatically assigned their Internet addresses when they boot up.

Chapter 11 Sharing Resources with Samba Explains how to share a Linux attached printer or CD-ROM drive with other Windows users. not upload. Uses a sample network of a few PCs. This chapter gives an explanation of the most popular Linux-based firewall software with step-by-step examples of a script/program you can write to protect your network. Chapter 15 Linux FTP Server Setup Shows how to set up a Linux server to allow uploading and downloading of files from the Internet using the File Transfer Protocol (FTP) package. of how to grant regular trusted users with temporary system administrator privileges and track what they do. with examples. This makes it easier to determine precisely who did what on a system. Linux. which is the typical scenario in schools. Chapter 10 Windows. configure. Chapter 13 Linux Wireless Networking Coverage of how to install. which can be virtually impossible to determine if everyone shares the administrator password. files for the sake of security. and a sample iptables script is provided. FTP can be tricky to configure with firewalls. There is also a tutorial on how to set up an FTP server to allow people to download only. and test a wireless network that includes Linux servers. and homes. Not all the features—just enough so that a user will be comfortable in creating one if software installation requires it.00Harrison_FM. corporate departments. . Chapter 12 Samba Security and Troubleshooting Simple Samba problem resolution outlined in a logical step-by-step fashion. and Samba How to make your Linux server act transparently as a Windows file server using the Samba package. PART 2 The Linux Web Site Project Chapter 14 Linux Firewalls Using iptables Security is an essential part of Web life. Also provides explanations.qxd 2/25/05 12:39 PM Page xxxv Preface xxxv Chapter 9 Linux Users and sudo The basics on how to create Linux user accounts.

This chapter explains how to create a Web site simply by tacking on easy-tounderstand customizations to the end of the configuration file. This chapter covers how to set up a secure shell (SSH) server for encrypted communications as an alternative to Telnet. Chapter 20 The Apache Web Server After DNS is configured you need to set up a server to handle your Web site’s pages. Also offers an example of how to save the configuration files of various Cisco networking devices to a Linux server using TFTP. . Chapter 21 Configuring Linux Mail Servers Covers how to configure the sendmail mail server to manage mail for your Web site. sendmail can be difficult to configure. Also covers how to configure SSH to allow you to log in to a remote Linux server without a password and to execute commands on a remote Linux server on demand. Regular DNS relies on an Internet network whose Internet address doesn’t change. one of the Internet’s greatest nuisances. Chapter 19 Dynamic DNS In many cases you might want to host a Web site on a DSL line. but the configuration file can be difficult to understand. This chapter covers how to track this constant change so that your Web site’s name (URL) always points to your new Internet address. Many home and office networks get their Internet addresses dynamically assigned.qxd 2/25/05 12:39 PM Page xxxvi xxxvi Preface Chapter 16 Telnet. which is all explained in this chapter. Full examples provided.00Harrison_FM. TFTP. which is also explained. Chapter 18 Configuring DNS How to configure the Domain Name Services (DNS) on a Linux server that will enable you to map an Internet address to a Web site name. Explains the different types of DNS servers and discusses simple methods of how to make them serve your departmental network. Chapter 17 Secure Remote Logins and File Copying Telnet is very insecure as the data passing between the client and server is unencrypted. The Apache Web server package is the most commonly used Linux software for this task. but it can be easily simplified using macros that are provided with it. are also explained. Both the Telnet and TFTP packages are governed by the xinetd program. and xinetd Coverage of how to log into a Linux server from a remote PC using the commonly available Telnet program. Methods on how to combat SPAM. as are the steps needed to retrieve your mail from your sendmail server.

which explains how to get graphs of CPU. as well as how to make your Linux server the primary time source for your network. and disk usage statistics with examples.qxd 2/25/05 12:39 PM Page xxxvii Preface xxxvii Chapter 22 Monitoring Server Performance Very detailed coverage on how to monitor your server’s performance. PART 3 Advanced Topics Chapter 25 Network-Based Linux Installation How to install Linux quickly over a network without the need for CDs. Chapter 24 The NTP Server Explanations on how to synchronize the time on your Linux server with wellknown time servers on the Internet. VMSTAT. . and even get Web-based Internet traffic graphs. Chapter 26 Linux Software RAID Coverage of how to configure Linux software to create redundant arrays of identical disk drives (RAID) to provide fail-safe data storage. Chapter 30 Configuring NIS A description of how to use name information services (NIS) to create a centralized username/password database for all your Linux servers. Chapter 28 Managing Disk Usage with Quotas Examples of how to limit the amount of disk space individual Linux users can use on a filesystem. This chapter explains how to do it. TOP. Chapter 23 Advanced MRTG for Linux Expanded coverage of MRTG. MRTG. and FREE utilities. memory. using the SNMP. with examples. Chapter 27 Expanding Disk Capacity Sometimes you run out of space on a Linux partition and need to expand it onto another hard disk. Chapter 29 Remote Disk Access with NFS A practical guide on how to use the network file system (NFS) to share files between Linux systems.00Harrison_FM. Webalizer. Also covers how to automate installation for large-scale deployments.

Appendix II Codes. yet often overlooked. such as accounting packages and message boards. The lightweight directory access protocol (LDAP) can also be used to store username/password data in a rapid access database format over an encrypted network connection. and Configurations Provides the library of useful sample configuration files and productivity enhancing scripts that are used as references in the main text. topics for Linux servers. . Chapter 34 Basic MySQL Configuration Many add-on Linux applications require interaction with a database of some sort.qxd 2/25/05 12:39 PM Page xxxviii xxxviii Preface Chapter 31 Centralized Logins Using LDAP and RADIUS NIS has its limitations because it is geared toward Linux logins and does unencrypted authentication.00Harrison_FM. a description of Security Enhanced Linux. that don’t run Linux. Chapter 33 Modifying the Kernel to Improve Performance A brief explanation with examples on how to configure the Linux kernel to improve system performance. This was included because many Linux administrators also have a networking and security management role as well. Appendix IV syslog Configuration and Cisco Devices Shows how to configure many popular Cisco network devices to send their status and error messages to a Linux server. This chapter explains how. When used in conjunction with RADIUS. adjusting dynamic kernel parameters. and how to manage your Linux server via a connection to the COM port. including additional security using TCP wrappers. Appendix I Miscellaneous Linux Topics Covers a variety of useful. Scripts. it can be used to authenticate users logging into a variety of devices. Appendix III Differences Between The Fedora Versions Describes many of the subtle differences in the functioning and configuration of the various versions of Fedora and Red Hat Linux. Also covers how to provide password-only access to the Internet. such as networking equipment. This is a handy brief reference when you are considering an upgrade of your operating system. Chapter 35 Configuring Linux VPNs Details of how to create a permanent site-to-site VPN between two offices. one of the most popular is MySQL. Chapter 32 Controlling Web Access with Squid This chapter focuses on how to use Squid to limit the amount of time users can spend browsing the Internet using the Squid Linux package. This chapter discusses basic MySQL configuration for the purposes of supporting the installation of third-party Linux applications.

00Harrison_FM.qxd 2/25/05 12:39 PM Page xxxix .

One of them. Christy Hackerd. Also. and George Vish. But the real dessert was reading the reviews of (alphabetically listed) Joseph Brazeal. too many to completely acknowledge on a single page. with Brenda Mulligan. Thank you all. I must first thank my family—Diana. Jack.qxd 2/25/05 12:39 PM Page xl Acknowledgments There have been numerous people who have provided me with the inspiration to write. Ever since that simple start. . without the combined knowledge of my coworkers the task would have been nearly impossible. Nancy Albright.00Harrison_FM. visitors to the site have sent numerous e-mails of encouragement. Blaze. who challenged me to improve the technical content and readability of the book. Barbara. Without them all the cake would have no icing. Robert Husted. who had the vision to take a chance. and finding omissions. suggesting new directions. Pauline. and Helene—who provided encouragement since the book was a single Web page on wireless Linux. correcting errors. Joseph Naberhaus. Bret Strong. provided me with the extra push to seek publication. Ron McCarty. Keith Lofstrom. It would also be unthinkable not to mention Jill Harry of Prentice Hall. Jade. and Linda Laflamme who always directed me along the right path.

the techie. and was the principal systems engineer responsible for the computerization of the island’s tax collection and social security systems. Jamaica’s first computer user group. Extensive use of Linux in this environment combined with his varied business background has helped him create this highly readable book for the newbie.qxd 2/25/05 12:39 PM Page xli About the Author Peter Harrison has been fascinated by computers since the early 1980s.00Harrison_FM. Long rides on his bicycle provide another guilty pleasure. He was the founding president of PCJAM. Before moving to Silicon Valley he ran Trinidad and Tobago’s first industrial trade office to Latin America. In his quieter moments. Peter enjoys the art and literature of the Caribbean and Latin America. and their bosses. He then sought new opportunities as the western Caribbean representative for a Fortune 500 pharmaceuticals firm and later became the international sales manager for a West Indian rum company. Peter likes to relax with his family on short weekend trips to the many attractions of the San Francisco Bay Area. Peter has since worked extensively in the Internet sector deploying large-scale data centers and Web sites. .

00Harrison_FM.qxd 2/25/05 12:39 PM Page xlii .

qxd 2/25/05 9:47 AM Page 1 P A R T I The Linux File Server Project 1 .01Harrison_Pt1.

01Harrison_Pt1.qxd 2/25/05 9:47 AM Page 2 .

People use them to create Web logs of their daily lives. Many companies rely almost exclusively on their Web sites to sell their products and provide both customer and supplier support services. provide family members with a place to store their memories. the book is divided into three sections of gradually increasing complexity to make this process easier.linuxhomenetworking. This book is about setting up Linux servers to do the things that most businesses and homes need. but later expanded them to become an important part of their operations. You have to consider factors of cost and convenience as well as service and support.02Harrison_ch01. It’s about getting the job done. Businesses originally used Web sites primarily as a marketing tool. The following is a typical Web site address: www. This chapter briefly addresses the most common issues and outlines the simple network architecture for use in a small office or home on which the focus of the rest of the book will be based. Not all businesses departments and homes require a Web site. but the process of establishing one touches many aspects of not only Linux. After this chapter. With this in mind. The decision as to whether to host your own Web site can be difficult. or tell people of their experiences in getting things to work.com. but information technology as well.qxd 2/25/05 9:46 AM Page 3 C H A P T E R 1 Why Host Your Own Site? IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ Our Network Alternatives to In-House Web Hosting Factors to Consider Before Hosting Yourself How to Migrate From an External Provider In-House Server Considerations Conclusion Web sites have proliferated greatly over the years to become a part of everyday life for many people. the first 3 .

02Harrison_ch01. and monitor your own Linux-based Web site on this network using a simple DSL or cable modem Internet connection.1 Wireless home network topology. but before you do it would be best to weigh the pros and cons of doing this. manage. OUR NETWORK The typical small office or home network is usually quite simple. the rest of the book shows how you can make a simple layout such as this become a functional low-volume Web site. There are companies that host multiple Web sites on the same server and are called virtual hosting providers. with a router/firewall connected to a broadband Internet connection protecting a single network on which all servers and PCs are connected.1.168.158. as seen in Figure 1. and troubleshooting before the first major project of using Linux as a main departmental file server for Windows PCs.1. The next section expands upon this knowledge to show you how to create.253. Wireless Home Network Topology Public Network Private Network ISP Gateway's IP Address: 97.100 Inside IP Address: 192.253.1 Wireless Card Workstation DHCP Figure 1. As stated before. Finally. the third section covers more advanced topics that will become invaluable as your Linux administration role expands.158. ALTERNATIVES TO IN-HOUSE WEB HOSTING There are two broad categories of hosting options for small Web sites.168.qxd 2/25/05 9:46 AM Page 4 4 Why Host Your Own Site? Chapter 1 section introduces you to networking.1. these are called dedicated hosting .30 Outside IP Address: DHCP or 97. There are also those that allow you to use servers completely dedicated to your site. software installation.25 (Depends on Tutorial) Wireless Card Internet Internet User ISP Gateway Firewall/Router WAP Linux Box (BIG BOY) 192.

the amount of power it consumes. Upload your Web pages to your private virtual hosting directory. you typically have to make contact with a live sales representative who represents an Internet data center. 4. and the amount of Internet bandwidth you expect to use. They will provide you with a user account.02Harrison_ch01. Dedicated Hosting In this scenario. from which to choose. though it is cheap. or they might offer a menu of services. which allows you to have a shopping cart and customer loyalty programs. such as my-web-site. with companies like Register. so you’ll probably have to wait at least that long before you’ll be able to view your site using your sub-domain. a number of e-mail addresses and an easy-to-use Web-based GUI to manage your settings. A 1 1⁄2-inch slot in a computer cabinet for a Web server can easily cost $200 per month for . you often have no control over the operation of the server and have to rely on the staff and operational procedures of the hosting company to get your changes implemented. Additional services such as backups. on-call engineering time.com. 2. monitoring. As you can imagine. these services can be fairly expensive. The steps are fairly straightforward: 1. www. Virtual Hosting It is easy to find virtual hosting companies on the Web that offer to host a simple Web site for about $10 per month.org.my-web-site. 3. the IP address of your site. The disadvantage of virtual hosting is that.com. firewall management. For an additional charge. It takes about 3-4 days for DNS to propagate across the Web. Register your domain name. and bandwidth graphing information often can be purchased as extra line items on your bill. many will also provide an e-commerce feature. from monitoring to backups. Sign up for the virtual hosting service.qxd 2/25/05 9:46 AM Page 5 Alternatives to In-House Web Hosting 5 providers.org. These may not necessarily be to your liking. Start testing your site using your IP address in your Web browser. Dedicated providers might provide you with only a network connection for a server you purchase and install in their data center. The virtual hosting provider will also offer free backups of your site. You must make sure your new domain name’s DNS records point to the DNS server of the virtual hosting company. Verisign. At a minimum. you have to pay for the amount of space your server occupies in the data center. technical support. or RegisterFree. and the name of a private directory on a shared Web server in which you’ll place your Web pages.

not those of the hosting provider. FACTORS TO CONSIDER BEFORE HOSTING YOURSELF Hosting your Web site externally. The advantage over virtual hosting is that you can customize the server to your needs.1 Factors to Consider When Hosting In-House Costs New hardware and software Risks Likelihood of a failure and its expected duration The cost of both the failure and post-failure recovery (hardware. you may also want to consider doing it yourself. time) Irregular procedures that could increase the vulnerability of your site to failure Savings Monthly outsourced Web hosting fee Elimination of the cost of delays to implement desired services Possible new application development Training The percentage of IT staff’s time installing and maintaining the site Potential cost of the risks (% likelihood of failure per month × cost of failure) Is In-House Hosting Preferred? There are a number of advantages and disadvantages to hosting Web sites that are physically under your own control. not what the ISP dictates.02Harrison_ch01. You won’t have to wait before changes are made and you can select the IT solution that best meets your needs. The pros and cons of doing this are listed in Table 1.qxd 2/25/05 9:46 AM Page 6 6 Why Host Your Own Site? Chapter 1 1Mbps of bandwidth. .1. Pros ☞ Increased Control and Flexibility: You will be able to manage all aspects of your Web site if it is hosted on a server based either in-house or within your control at a remote data center. Table 1. You can install the software you need. is the ideal solution for many small Web sites but there are a number of reasons why you may want to move your Web site to your home or small office. software. especially virtual hosting. data restoration. Despite the relative merits of external hosting.

DNS. If your home already has DSL there would be no additional network connectivity costs.qxd 2/25/05 9:46 AM Page 7 Factors to Consider Before Hosting Yourself 7 ☞ ☞ ☞ ☞ There is also the possibility of offsetting the cost of your server by subleasing space on it to other companies in your community. ☞ Security: One important factor to consider is the security of your new server. If you can’t get adequate support for billing. so that you can become a small virtual hosting service yourself. there are other costs to consider.02Harrison_ch01. and/or you may not have access to the foreign currency to host your site abroad. database services. and customer care services. You may also have to incur additional training costs to ensure that your IT staff has the necessary knowledge to do the job internally. The cost of using an external Web hosting provider will increase as you purchase more systems administration services. These services may be more difficult to implement in-house. Cons ☞ Lost Services: You lose the convenience of many of the services—such as backups. redundant hardware. load balancing. an in-house solution may be better. security audits. New Skills: An additional benefit is learning the new skills required to set up the site. You may not have access to cheap real estate in which to host your servers. Cost: It is possible to host a Web site on most DSL connections. Proceed with the server migration only if you feel your staff can handle the job. ☞ Cost: Though you may be able to save money on a data circuit. You will eventually be able to justify hosting your Web site in-house based on this financial fact. Always weigh the degree of security maintained by your hosting provider against the security you expect to provide in-house. Changes can be made with little delay. Availability: Reliable virtual hosting facilities may not be available in your country. Hosting providers may provide software patches to fix security vulnerabilities on your Web servers and may even provide a firewall to protect it. and technical support—offered by the virtual hosting company. So for a savings of $10 per month the project should pay for itself in less than a year. ☞ Technical Ability: Your service provider may have more expertise in setting up your site than you do. Commercial office space is often more expen- . engineering. ☞ Availability: In many cases the reliability of a data center’s Internet connectivity is usually higher than that of your broadband connection. A Web site can be hosted on this data circuit for only the additional hardware cost of a network switch and a Web server. Language: ISPs often provide technical support in only a few languages. You should be able to buy this equipment secondhand for about $100.

HOW TO MIGRATE FROM AN EXTERNAL PROVIDER Chapter 18. high-speed redundant SCSI disks that can be replaced while the system is running without affecting performance. Label all your cables at both ends and try to create an updated network diagram that you can show anyone who will provide you assistance. monitoring systems. I have one server . but you pay for the peace of mind when your only source of income is your Web site. receives more than half a million page views per month and runs on a 1GHz Intel Celeron with 1GB of RAM. backup systems. has a detailed explanation of the steps involved in migrating your Web site from an external hosting provider to your home or small office. A secondhand PC is adequate in this case. Wireless technology for a home-based Web site can be extremely convenient. such as UPSs. www. special error-correcting ECC RAM. maintenance contracts. which was the inspiration for this book.linuxhomenetworking. Try to have a dedicated area for your server that’s clean. Some companies use one color for networks using private IP addresses and another for Internet-facing networks. others use one color for straight-through and another for crossover cables. multiple PCI buses. Purpose-built Web servers tend to use multiple CPUs. special built-in diagnostic tools.com. You should also read Chapters 20. and slim-line cases only a few inches high.” and 21 “Configuring Linux Mail Servers.” which covers DNS. and additional power feeds. all of which may be already bundled in with the services of an external data center. “Configuring DNS. dual redundant power supplies. cool. You may have to purchase additional equipment and services to support your servers. You can place your small wireless router near your DSL/cable modem and the server anywhere in the house. “The Apache Web Server. IN-HOUSE SERVER CONSIDERATIONS For small Web sites without a great deal of database activity and where “hot standby” hardware isn’t a great need. Another good idea is to color-code your cables. In my little lab. a basic desktop system will work fine. and dry.02Harrison_ch01.” on Web and mail server configuration to help provide a more rounded understanding of the steps involved.qxd 2/25/05 9:46 AM Page 8 8 Why Host Your Own Site? Chapter 1 sive than basic data center space. and that uses UPS-protected power. They cost significantly more. software patch management.

DSL and wireless services are sometimes asymetrical.qxd 2/25/05 9:46 AM Page 9 In-House Server Considerations 9 behind a bookcase. You should be most concerned about upstream speed for your Web site to the Internet. This can have an impact on your Web site preparation and will be discussed in more detail in later chapters. Internet service providers (ISPs) provide asymmetric services for residential users and the downstream rate is almost always higher than the upstream. Red Hat Linux was a free product. A popular one is Windows. Once Fedora updates are proven . one under a bed. Some technologies. but the service also has a per kilometer monthly distance charge. but the risk is that a book falling behind a bookcase or a bounce from a vacuum cleaner could take your site off the air. and provide better customer support. such as T1 links. can optionally provide a dedicated circuit between two locations external to the Internet. offer more bandwidth. This book focuses on Fedora Linux with some references to Red Hat Linux. there may be no other choice. At the same time Red Hat decided to create Fedora Linux as a support-free product with an aggressive development cycle. residential customers usually get a dynamic address allocation. What’s the difference? Until Version 9. its popular corporate cousin. The ISP will usually provide the business with a fixed range of Internet addresses as part of the service. which also has a lower overall total cost of ownership.02Harrison_ch01. They reserve symmetrical data circuits for businesses that usually need high bandwidth to both surf the Web and serve Web pages. in that the incoming downstream data rate from the Internet is different than the reverse outgoing upstream speed. Another source of concern would be deciding on the operating system to use. especially for smaller companies. another behind the TV. New versions of Fedora are released every six months. Inbound Web browser queries don’t use a lot of data bandwidth. and wireless links. but the Web pages that contain the outbound replies do. The company then decided to create enterprise and desktop versions that had paid service contracts bundled with them. DSL. These issues are becoming less important as software vendors are increasingly porting their applications to Linux. be more reliable. an increasingly strong rival to Windows. but they may not be available in your area or the installation times may not be acceptable. and a couple around my desk. Though the original applications may be developed by volunteers. which may be the only product your Web or business application will work with and with which your staff is most familiar. which is generally unsuitable for businesses that often require more stability and support. the Fedora versions are maintained by Red Hat. and these maintained the Red Hat brand. When you live in an apartment. High-speed links are usually marketed to businesses and their cost per megabyte of data transfer is usually higher because the service may be combined with data center space. There are many data circuit technologies such as cable modem. Selecting an Internet connection for your Web site is not as easy as it seems.

CONCLUSION The decision to manage your Web site in-house can be difficult. This may not be the one suitable for you.qxd 2/25/05 9:46 AM Page 10 10 Why Host Your Own Site? Chapter 1 stable. this book will provide a lot of guidance in completing a successful project. and other Linux distributions should also be considered. You don’t have to get a purchase order to play with Fedora. and management. Constant communication between Red Hat and the developers help to keep the updates synchronized. plan carefully.02Harrison_ch01. . and always be aware of the potential risks of the decision you make. your peers. When you become comfortable with it and have proven the concept to yourself. I chose Fedora because it’s free. Whatever you choose to do. which are updated every 12–18 months. If you decide to do it. even if it’s informal. I also chose Fedora Linux because it’s popular and it’s the Linux flavor I’ve worked with most frequently at home and at work. you can then consider the more stable Red Hat equivalent. Always get a professional opinion. they are incorporated into the Red Hat Linux releases.

qxd 2/25/05 9:55 AM Page 11 C H A P T E R 2 Introduction to Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ The OSI Networking Model An Introduction to TCP/IP How IP Addresses Are Used to Access Network Devices How Subnet Masks Group IP Addresses Into Networks The Physical and Link Layers Networking Equipment Terminology Additional Introductory Topics Conclusion Installing the Linux operating system is only the first step toward creating a fully functional departmental server or Web site. general software installation. 11 . It will help make the road to Linux mastery less perilous. Almost all computers are now networked in some way to other devices. therefore. These chapters will then introduce the remaining chapters that cover Linux troubleshooting. and even you. Familiarity with the concepts explained in the following sections will help answer many of the daily questions often posed by coworkers.03Harrison_ch02. a road that begins with an understanding of the OSI networking model and TCP/IP. a basic understanding of networking and issues related to the topic will be essential to feeling comfortable with Linux servers. friends. This introductory chapter forms the foundation on which the following network configuration and troubleshooting chapters will be built. and the configuration of many of the most popular Linux applications used in corporate departments and Small Office/Home Office (SOHO) environments.

The very first layer of the OSI model describes the transmission attributes of the cabling or wireless frequencies used at each link or step along the way.qxd 2/25/05 9:55 AM Page 12 12 Introduction to Networking Chapter 2 THE OSI NETWORKING MODEL The Open System Interconnection (OSI) model. In the OSI model each component along the data communications path is assigned a layer of responsibility—in other words. our friend has to hand over a permit to enter the country. Converts data from one presentation format to another—for example. she asks the border guards for directions to reach the next frontier and displays the permit for that new kingdom as proof that she has a legitimate reason for wanting to go there. layer 3 ensures that the data can hop from link to link on the way to the final destination described in its header. Once inside. Each layer extracts the permit. When the data finally arrives. A good analogy would be a traveler who prepares herself to return home through many dangerous kingdoms by obtaining permits to enter each country at the very beginning of the trip. This layer also strips away its permit and forwards the data to the next layer. it needs from the data and uses this information to correctly forward what’s left to the next layer.03Harrison_ch02. and so the cycle continues for seven layers. A description of the functions of each layer in the model can be seen in Table 2. e-mail text entered into Outlook Express being converted into SMTP mail formatted data.1. At each frontier.1 Layer 7 6 The Seven OSI Layers Name Application Presentation Description The user interface to the application. the layer 4 header is used to determine which locally installed software application should receive it. a kingdom over which it rules. developed by the International Organization for Standardization. or header information. layer 7 defines what the end user will see in the form of an interface. Table 2. graphical on a screen or otherwise. Manages continuing requests and responses between the applications at both ends over the various established connections. Application telnet FTP sendmail 5 Session . Layer 2 describes the error correction methodologies to be used on the link. defines how the various hardware and software components involved in data communication should interact with each other. The application uses the guidelines of layer 5 to keep track of the various communications sessions it has with remote computers and uses layer 6 to verify that the communication or file format is correct. Finally.

It is part of the larger OSI model upon which most data communications is based. Defines the electrical and physical characteristics of the network cabling and interfacing hardware. It then passes the data to the program that expects to receive it. the resulting data is said to be a packet. After the link’s overhead bytes have been stripped away.03Harrison_ch02. each with its own error detection bytes in the control section or header of the packet. One component of TCP/IP is the Internet Protocol (IP). the data is usually split into multiple pieces or packets. For manageability. Application TCP UDP IP ARP Ethernet ARP Ethernet AN INTRODUCTION TO TCP/IP TCP/IP is a universal standard suite of protocols used to provide connectivity between networked devices. After the packet’s overhead bytes have been stripped away. Error control and timing of bits speeding down the wire between two directly connected devices. The remote computer then receives the packets and reassembles the data and checks for errors.qxd 2/25/05 9:55 AM Page 13 An Introduction to TCP/IP 13 Layer 4 Name Transport 3 Network 2 Link 1 Physical Description Manages the establishment and tearing down of a connection. Data sent on a link is said to be structured in frames. which is responsible for ensuring that data is transferred between two addresses without being corrupted. Handles the routing of data between links that are not physically connected. Ensures that unacknowledged data is retransmitted. the resulting data is said to be a segment. Correctly resequences data packets that arrive in the wrong order. .

TCP acts in a similar way. SYN-ACK. to which the other server also replies with a FIN-ACK segment. TCP keeps track of the packets sent by giving each one a sequence number with the remote server sending back acknowledgment packets confirming correct delivery. This informs the computer receiving the data about the type of layer 4 transportation mechanism being used. which is used to determine which network application on the computer should process the data. Someone knocks on the door to a house. The two most popular transportation mechanisms used on the Internet are Transmission Control Protocol (TCP) and User Datagram Protocol (UDP).” The conversation ends with a final “bye” from the hungry person.03Harrison_ch02. This SYN. each with only the ACK bit set. “I think I’m finished here too. the TCP/UDP header is then inspected for the port value. Both persons knew who was on the other side of the door before it opened and a conversation can now begin.” to which the reply is. The server initiating the connection sends a segment with the SYN bit set in TCP header. . to which the visitor replies. TCP is a good example of a connection-oriented protocol. When the type of transport protocol has been determined. This is explained in more detail later. “Who is it?”. “It’s me!” Then the door opens.qxd 2/25/05 9:55 AM Page 14 14 Introduction to Networking Chapter 2 How does the computer know what program needs the data? Each IP packet also contains a piece of information in its header called the type field. it sends a segment to the other with the FIN and ACK bits set. see you tomorrow. the person inside asks. When one of the servers needs to end the communication. The communication terminates with a final ACK from the server that wanted to end the session. I have to go for lunch. How TCP Establishes a Connection Any form of communication requires some form of acknowledgment for it to become meaningful. The target replies with a segment with the SYN and ACK bits set. to which the originating server replies with a segment with only the ACK bit set. Programs that use TCP therefore have a means of detecting connection failures and requesting the retransmission of missing packets.” The communication then continues with a series of segment exchanges. This is the equivalent of ending a conversation by saying “I really have to go now. TCP Is a Connection-Oriented Protocol TCP opens up a virtual connection between the client and server programs running on separate computers so that multiple and/or sporadic streams of data can be sent over an indefinite period of time between them. ACK mechanism is often called the “three-way handshake.

etc. is the sequential serial number of the first byte of the next segment it expects to receive from the other end. Data is sent on a “best effort” basis with the machine that sends the data having no means of verifying whether the data was correctly received by the remote machine.03Harrison_ch02. ACK] Seq=8404 Ack=9767 Win=5792 -> -> -> -> -> -> hostB hostB hostA hostA hostB hostA TCP 1443 > http [ACK] HTTP HEAD / HTTP/1. UDP. The . not to be confused with the ACK bit. but both the UDP and the TCP segment headers track the port being used. TCP’s “Connectionless” Cousin UDP is a connectionless protocol.qxd 2/25/05 9:55 AM Page 15 An Introduction to TCP/IP 15 Here is a modified packet trace obtained from the tethereal program discussed in Chapter 4. UDP is usually used for applications in which the data sent is not mission-critical. and the total number of bytes cannot exceed the Win or window value that follows it. Seq=9767 Ack=8405 Win=5840 Len=0 Seq=8405 Ack=9985 Win=54 Len=0 Seq=9985 Ack=8672 Win=6432 Len=0 ACK] Seq=8672 Ack=9985 Win=54 -> hostB TCP 1443 > http [FIN. the receiver will re-send the requesting segment asking for the information to be sent again.1 TCP http > 1443 [ACK] HTTP HTTP/1. It is also used when data needs to be broadcast to all available servers on a locally attached network where the creation of dozens of TCP connections for a short burst of data is considered resource-hungry. The TCP code keeps track of all this along with the source and destination ports and IP addresses to ensure that each unique connection is serviced correctly. a random value of 9766 was assigned to the first byte. and all subsequent bytes for the connection from this host will be sequentially tracked. TCP and UDP Ports The data portion of the IP packet contains a TCP or UDP segment sandwiched inside.1 200 OK TCP 1443 > http [ACK] TCP http > 1443 [FIN. ACK] Seq=9985 Ack=8673 Win=6432 -> hostA TCP http > 1443 [ACK] Seq=8673 Ack=9986 Win=54 In this trace. the sequence number represents the serial number of the first byte of data in the segment. The acknowledgment number. Only the TCP segment header contains sequence information. the third number 9768. “Simple Network Troubleshooting.” You can clearly see the three-way handshake to connect and disconnect the session: hostA hostB Len=0 hostA hostA hostB hostB hostA hostB Len=0 hostA Len=0 hostB -> hostB TCP 1443 > http [SYN] Seq=9766 Ack=0 Win=5840 Len=0 -> hostA TCP http > 1443 [SYN. This makes the second byte in the segment number 9767. So in the first line. If data isn’t received correctly. or Ack.

It is not strictly a TCP/IP protocol. control. say. IP packets will occasionally arrive .03Harrison_ch02. Remember this concept because it will be helpful in understanding the traceroute troubleshooting technique outlined in Chapter 4 that covers network troubleshooting. and port 25 is reserved for SMTP e-mail. If the TTL value reaches 0. 2049 and query the server on port 80 (HTTP). The server sending the packet sets the initial TTL value. and those above 1024 are generally reserved for non-system third-party applications. Usually when a connection is made from a client computer requesting data to the server that contains the data: ☞ The client selects a random previously unused source port greater than 1024 and queries the server on the destination port specific to the application. This mechanism helps to ensure that bad routing on the Internet won’t cause packets to aimlessly loop around the network without being removed. ☞ The server recognizes the port 80 request as an HTTP request and passes on the data to be handled by the Web server software. When the Web server software replies to the client. The TCP/IP Time to Live Feature Each IP packet has a Time to Live (TTL) section that keeps track of the number of network devices the packet has passed through to reach its destination. the client will use a source port of. Certain programs are assigned specific ports that are internationally recognized.” but a response to the initial port 80 HTTP query. If it is an HTTP request. For example. port 80 is reserved for HTTP Web traffic. it tells the TCP application to respond back to port 2049 of the client using a source port of port 80. Ports less than or equal to 1024 are reserved for privileged system functions. The ICMP Protocol and Its Relationship to TCP/IP There is another commonly used protocol called the Internet Control Message Protocol (ICMP). For example. and each network device that the packet passes through then reduces this value by 1. but TCP/IP-based applications use it frequently. the network device will discard the packet. ☞ The client keeps track of all its requests to the server’s IP address and will recognize that the reply on port 2049 isn’t a request initiation for “NFS.qxd 2/25/05 9:55 AM Page 16 16 Introduction to Networking Chapter 2 source/destination port and the source/destination IP addresses of the client and server computers are then combined to uniquely identify each data flow. and informational messages for use by the operating system. ICMP provides a suite of error. TTLs therefore help reduce the clogging of data circuits with unnecessary traffic.

” which covers Linux specific networking topics. The Internet Assigned Numbers Authority (IANA) is the organization responsible for assigning IP addresses to Internet Service Providers (ISPs) and deciding which ones should be used for the public Internet and which ones should be used on private networks. ICMP also includes echo and echo reply messages used by the Linux ping command to confirm network connectivity. HOW IP ADDRESSES ARE USED TO ACCESS NETWORK DEVICES All TCP/IP-enabled devices connected to the Internet have an Internet Protocol (IP) address.0. Private IP Addresses Some groups of IP addresses are reserved for use only in private networks and are not routed over the Internet.255 Home networking equipment/devices usually are configured in the factory with an IP address in the range 192. “Linux Networking.168.255.0 . IP addresses are in reality a string of 32 binary digits or bits. network engineers often divide these 32 bits into four sets of 8 bits (or octets).10.1.qxd 2/25/05 9:55 AM Page 17 How IP Addresses Are Used to Access Network Devices 17 at a server with corrupted data due to any number of reasons including a bad connection. More information on ICMP messages can be found in both Appendix II.168. each representing a number from 0 to 255.0. or even misconfiguration.1 to 192.65.172.1. Just like a telephone number. “Codes. These are called private IP addresses and have the following ranges: 10. It will then issue an ICMP reject message to the original sending machine saying that the data should be re-sent because the original transmission was corrupted. ICMP TTL expired messages are also sent by network devices back to the originating server whenever the TTL in a packet is decremented to 0. You may be wondering how devices using private addresses could ever access the Internet if the use of private addresses on the Internet is illegal.25.) to create the familiar dotted decimal notation.0.16. and Configurations.255. .03Harrison_ch02. The server will usually detect this by examining the packet and correlating the contents to what it finds in the IP header’s error-control section.” and Chapter 4. Note Chapter 3. Scripts.255 172.168.12. For ease of use. electrical interference. explains how to configure the IP address of your Linux box. it helps uniquely identify a user of the system. Each number is then separated by a period (.192.0 .255.255.0.255.255 192.168. An example of an IP address that follows these rules is 97.31.0 .

You can also use many to one NAT.0. You can configure NAT to be one to one. Network Address Translation (NAT) Makes Private IPs Public Your router/firewall will frequently be configured to give the impression to other devices on the Internet that all the servers on your home/office network have a valid public IP address and not a private IP address. This is called network address translation (NAT) and is often also called IP masquerading in the Linux world. Basic NAT testing requires you to ask a friend to try to connect to your home network from the Internet.qxd 2/25/05 9:55 AM Page 18 18 Introduction to Networking Chapter 2 The situation gets even more confusing if you consider the fact that hundreds of thousands of office and home networks use these same addresses. ☞ Hundreds of PCs and servers behind a NAT device can masquerade as a single public IP address. NAT protects your home PCs by assigning them IP addresses from private IP address space that cannot be routed over the Internet. The localhost IP Address Whether or not your computer has a network interface card it will have a built-in IP address with which network-aware applications can communicate with one another. “Linux Firewalls Using iptables. . but they will be explained in later sections of this chapter.03Harrison_ch02. This must cause networking confusion. Some of the terms mentioned here might be unfamiliar to you.” that covers the Linux iptables firewall. As a general rule. and will be revisited in many later chapters. you won’t be able to access the public NAT IP addresses from servers on your home network. in which you request your ISP to assign you a number of public IP addresses to be used by the Internetfacing interface of your firewall. There are many good reasons for this. and then you pair each of these addresses to a corresponding server on your protected private IP network.0. Don’t worry. This prevents hackers from directly attacking your home systems because packets sent to the private IP will never pass over the Internet.1 and is frequently referred to as localhost. the two most commonly stated are: ☞ No one on the Internet knows your true IP address. This concept is important to understand. This IP address is defined as 127. Examples of NAT may be found in the IP masquerade section of Chapter 14. this problem is overcome by NAT. This greatly increases the number of devices that can access the Internet without running out of public IP addresses. in which the firewall maps a single IP address to multiple servers on the network.

all servers accessing the Internet will appear to have the single public IP address of the router/firewall because of many to one NAT. port forwarding is one of the most common methods used to host Web sites at home with DHCP DSL. For example. subnet mask default routers.03Harrison_ch02. at which point the DHCP client and server renegotiate a new IP address from the server’s predefined pool of addresses. subnet mask. Because the router/firewall is located at the border crossing to the Internet. you can configure your router/firewall to forward TCP port 80 (Web/HTTP) traffic destined to the outside NAT IP to be automatically relayed to a specific server on the inside home network As you may have guessed. and other IP parameters. Port forwarding is a method of counteracting this. DNS server. The reverse isn’t true. and default gateway information. The DHCP client sends out a query requesting a response from a DHCP server on the locally attached network. The DHCP server then replies to the client PC with its assigned IP address. it can easily keep track of all the various outbound connections to the Internet by: ☞ Monitoring the IP addresses and TCP ports used by each home-based server ☞ Mapping it to the TCP ports and IP addresses of the Internet servers with which they want to communicate This arrangement works well with a single NAT IP trying to initiate connections to many Internet addresses. New connections initiated from the Internet to the public IP address of the router/firewall face a problem. The assignment usually occurs when the DHCP-configured machine boots up or regains connectivity to the network. The router/firewall has no way of telling which of the many home PCs behind it should receive the relayed data because the mapping mentioned earlier doesn’t exist beforehand. . DHCP The Dynamic Host Configuration Protocol (DHCP) is a protocol that automates the assignment of IP addresses. In this case the data is usually discarded. The assignment of the IP address usually expires after a predetermined period of time. You’ll probably have to allow access for the entire remote DHCP subnet for a particular TCP/UDP port.qxd 2/25/05 9:55 AM Page 19 How IP Addresses Are Used to Access Network Devices 19 Port Forwarding with NAT Facilitates Home-Based Web Sites In a simple home network. Configuring firewall rules to accommodate access from machines that receive their IP addresses via DHCP is therefore more difficult because the remote IP address will vary from time to time.

“Configuring DNS. which then redirects the query to the linuxhomenetworking. It explains the auxiliary DNS standard called dynamic DNS (DDNS). Chapter 18. The linuxhomenetworking. . Each server in the chain will store the most frequent DNS name to IP address lookups in a memory cache that helps speed up the response. 4. Your ISP’s DNS server then probably redirects your query to one of the 13 root name servers. The most commonly used form of DSL will also assign the outside interface of your router/firewall with a single DHCP-provided public IP address. Most home computers get the IP address of their DNS server via DHCP from their router/firewall. 1. 6.” explains how to configure DNS for a Web site housed on a DHCP DSL circuit where the IP address constantly changes.” to find out how to make your Linux box provide the DHCP addresses for the other machines on your network. 2.linuxhomenetworking. You can also look at Chapter 8. this process can cause a noticeable delay when you are browsing the Web. You can also make your Linux box into a DHCP server.com into an IP address that can be used behind the scenes by your computer.qxd 2/25/05 9:55 AM Page 20 20 Introduction to Networking Chapter 2 Most home router/firewalls are configured in the factory to be DHCP servers for your home network.com domain’s name server.linuxhomenetworking. once it has a fixed IP address. The router/firewall then redirects the DNS queries from your computer to the DNS name server of your ISP.” explains how you can make your Linux box into a caching or regular DNS server for your network or Web site if your ISP provides you with fixed IP addresses. “Configuring the DHCP Server. Here is a step-by-step description of what happens with a DNS lookup. “Dynamic DNS. which was created for this type of scenario. 3. You can check Chapter 3 for how to configure your Linux box to get its IP address via DHCP. As you can imagine.com DNS name servers. 5. Chapter 19.03Harrison_ch02.com.com domain name server then responds with the IP address for www. The root server then redirects your query to one of the Internet’s . How DNS Links Your IP Address to Your Web Domain The domain name system (DNS) is a worldwide server network used to help translate easy-to-remember domain names like www. Home router/firewall providing DHCP services often provides its own IP address as the DNS name server address for home computers.

Expect it to become an increasingly bigger part of your network planning in years to come. Most home networks use a subnet mask of 255. A simple analogy would be a phone number. HOW SUBNET MASKS GROUP IP ADDRESSES INTO NETWORKS Subnet masks are used to tell which part of the IP address represents the network on which the computer is connected (network portion) and the computer’s unique identifier on that network (host portion).03Harrison_ch02. this book uses the latter term for the sake of consistency.1. such as video and voice. The term netmasks is often used interchangeably with the term subnet masks.168.1. which I have described here.168. IPv6 also inherently supports the IPSec protocol suite used in many forms of secured networks. billion times more IP addresses. The (808) represents the area code.0. billion. 192.25 and a subnet mask of 255.168. You should also be aware that there is now a version 6 (IPv6) that has recently been developed as a replacement. which it is hoped should last for most of the 21st century.qxd 2/25/05 9:55 AM Page 21 How Subnet Masks Group IP Addresses into Networks 21 IP Version 6 (IPv6) Most Internet-capable networking devices use version 4 of the Internet Protocol (IPv4). So if your server has an IP address of 192. This means you can then use IP addresses from #1 to #254 on your private network.1 and the server or host would be device #25 on that network. the allocation of version 4 addresses will soon be exhausted between all the world’s ISPs. such as virtual private networks (VPNs). Each 255 means this octet is for the area code (network portion). the network portion would be 192. and the 225-2468 represents the telephone within that area code. Version 6. Subnet masks enable you to specify how long you want the area code to be (network portion) at the expense of the number of telephones that are in the area code (host portion). over less time-sensitive ones. Most current operating systems support IPv6 even though it isn’t currently being used extensively within corporate or home environments.255.255. IPv6 packets are also labeled to provide quality-of-service information that can be used in prioritizing real-time applications.0 would be the network address and 192. In our example.1. . offers 80 billion. such as regular Web surfing and chat. the first IP address in a network is reserved as the network’s base address and the last one is reserved for broadcast traffic that is intended to be received by all devices on the network. In all cases. With only 32 bits.168.255.255 would be used for broadcasts.0. such as (808) 225-2468. which uses a much larger 128-bit address.255.

255.255.253.255.255.255. Calculating the Range of Addresses on Your Network If someone gives you an IP address of 97.255.qxd 2/25/05 9:55 AM Page 22 22 Introduction to Networking Chapter 2 Calculating the Number of Addresses Assigned to a Subnet Most office and home networks use networks with 255 IP addresses or less in which the subnet mask starts with the numbers 255.248.255.255. This gives you the theoretical number of subnets of the same size that are below this IP address. the boundaries of your network? The following section outlines the steps to do this using both a manual and programmed methodology. Table 2. 28 / 8 = 3).255.255. In many cases the subnet mask isn’t referred to by the dotted decimal notation. 192.252 So for example.255. so I’ll not discuss larger networks because that can become complicated. Here we go! 1.240 255.128 255. Subtract the last octet of the subnet mask from 256 to give the number of IP addresses in the subnet (256 – 248 = 8). but in cases where less than 255 IP addresses are required a few rules apply. 224. There are only seven possible values for the last octet of a subnet mask.255.255.255. 240.255.255. 128. but rather by the actual number of bits in the mask.248 255. These are 0.158. Divide the last octet of the IP address by the result of step 1. if you have a subnet mask of 255.28 and a subnet mask of 255.192.255.2. You can calculate the number of IP addresses for each of these by subtracting the value from 256.0 255. 2.224 255.255. 248 and 252.2 The Dotted Decimal and Slash Subnet Mask Notations Slash Format /24 /25 /26 /27 /28 /29 /30 Available Addresses 256 128 64 32 16 8 4 Dotted Decimal Format 255.03Harrison_ch02.192 255.255.255. This is not a pure networking text. A list of the most commonly used masks in the office or home environment is presented in Table 2. don’t bother with the remainder (for example.255. you have 64 IP addresses in your subnet (256 – 192). how do you determine the network address and the broadcast address—in other words. So for example a mask of 255.0 may be called a /24 mask instead. Manual Calculation Take out your pencil and paper—manual calculation can be tricky.255. .

168. 256 – 240 = 16 75 / 16 = 4 16 × 4 = 64. Therefore the broadcast address is 192. It will accept subnet masks in dotted decimal format or /value format: [root@bigboy tmp]# . Here is an example of how to use it.158.3. The broadcast address is 97.255. 4.168.32. this calculation exercise only works with subnet masks that start with “255. Let’s do this for 192. Note As a rule of thumb. Think of it as the broadcast address being the network address plus the number of IP addresses in the subnet minus 1. 64 + 16 –1 = 79.92 Network Base Address : 216. (24 + 8 –1 = 31).253.80 Broadcast Address : 216.193.0.0. 3.151.255.168.253.240: 1. 3. If you are sub-netting a large chunk of IP addresses it’s always a good idea to lay it out on a spreadsheet to make sure there are no overlapping subnets.224: 1.0.255. The network address is therefore 97. 4.0. 2.0.193. 2.193.255”. Let’s do this for 10. the last octet of your network base address must be divisible by the “256 minus the last octet of your subnet mask” and leave no remainder.151.151.0. Therefore the broadcast address is 10.240 : 16 IP Addresses .75 with a mask of 255.193. Multiply this result by the result of step 1 to get the network address (8 × 3 = 24). Therefore the network base address is 10. Calculation Using a Script There is a BASH script in Appendix II that will do this for you.31. 4.151.64.255. 32 + 32 – 1 = 63. just provide the IP address followed by the subnet mask as arguments.63.24.158.255.sh 216.79. Once again.3. Think of it as the third subnet with 8 addresses in it. Therefore the network base address is 192.95 Subnet Mask Subnet Size [root@bigboy tmp]# : 255.qxd 2/25/05 9:55 AM Page 23 How Subnet Masks Group IP Addresses into Networks 23 3.3.03Harrison_ch02.255. The broadcast address is the result of step 3 plus the result of step 1 minus 1./subnet-calc.92 /28 IP Address : 216.255. 256 – 224 = 32 56 / 32 = 1 32 × 1 = 32.56 with a mask of 255.

158. There are very many standards that define the physical. the server would assume that it transmitted data simultaneously with another server on the cable.31 Network base address Gateway Available Available Available Available Available Broadcast THE PHYSICAL AND LINK LAYERS TCP/IP can be quite interesting.253.158.25. or cyclic redundancy check (CRC). detected any differences between the two.qxd 2/25/05 9:55 AM Page 24 24 Introduction to Networking Chapter 2 Subnet Masks for the Typical Business DSL Line If you purchased a DSL service from your ISP that gives you fixed IP addresses.25 97.03Harrison_ch02.253. because without them. which was strung between multiple devices. Computers would wait until the line was clear before transmitting.29 97.158.253. if the ISP provides you with a public network address of 97. too. They would then send their data while comparing what they wanted to send with what they actually sent on the cable as a means of error detection. You will frequently see references to Ethernet being a CSMA/CD technology for this reason. . If a mathematical comparison.158. and error-control methodologies of data communication. your IP addresses will be 97.255. they will most likely provide you with a subnet mask of 255.253.248 that defines 8 IP addresses.158.253.27 97.253.158. also known as collisions. but the data transmission and error correction strategy is the same in all.24.158. but knowledge of the first two layers of the OSI model is important. is called collision detect or CD. and a gateway of 97.255. multiple access or CSMA. which is available in a variety of cable types and speed capabilities.255. One of the most popular ones in departmental networks is Ethernet. For example. The ability to detect garbling due to simultaneous data transmissions.253.24 97. even the most basic communication would be impossible.26 97.158. and similar schemes are now being used in wireless networks. a subnet mask of 255.253.158.248.253.255. Ethernet used to operate primarily in a mode where every computer on a network section shared the same Ethernet cable.28 97. It would then wait some random time and retransmit at some later stage when the line was clear again.30 97.158. electrical. had the correct signaling levels is a methodology called carrier sense.253. Transmitting data only after first sensing whether the cable.

NETWORKING EQUIPMENT TERMINOLOGY Up to this point you have had only an introduction to the theory of the first two OSI layers. . Full duplex uses separate pairs of wires for transmitting and receiving data so that incoming data flows don’t interfere with outgoing data flows.qxd 2/25/05 9:55 AM Page 25 Networking Equipment Terminology 25 Ethernet devices are now usually connected via a dedicated cable. Network Interface Cards Your network interface card is also frequently called a NIC. Error-detection and dataretransmission mechanisms ensure that the data reaches the destination correctly and are specifically designed to remedy data corruption caused when multiple devices start transmitting at the same time. and T1 circuits are all parts of these layers. The next few sections describe many physical and link layer concepts and the operation of the devices that use them to connect the computers in our offices and homes. Devices that want to transmit information have to wait their turn until the coast is clear. This indicates that you are using the correct type of cable and that the duplex has been negotiated correctly between the devices at both ends. cable modem standards. the most recent versions can handle up to 40Gbps! The 802. DSL. Now we’ll cover the hardware used to implement them. but in only one direction at a time. using more powerful hardware capable of simultaneously transmitting and receiving without interference. Duplex Explained Full duplex data paths have the capability of allowing the simultaneous sending and receiving of data. Currently.11 specifications that define many wireless networking technologies are another example of commonly used layer 1 and 2 components of the OSI model. the most common types of NIC used in the home and office are Ethernet and wireless Ethernet cards. Half duplex data paths can transmit in both directions too. thereby making it more reliable and inherently faster than its predecessor versions. The Meaning of the NIC Link Light The link light signifies that the NIC card has successfully detected a device on the other end of the cable. Half duplex uses the same pairs of wires for transmitting and receiving data.03Harrison_ch02. The original Ethernet standard has a speed of 10Mbps. at which point they send the data.

If there is no ARP entry. The server first checks its routing table to see which router provides the next hop to the destination network.1. When a reply is received.1. 4. Every IP packet is sent out of your NIC wrapped inside an Ethernet frame that uses MAC addresses to direct traffic on your locally attached network. As the packet hops across the Internet.168. If there is an ARP entry. How ARP Maps the MAC Address to Your IP Address The Address Resolution Protocol (ARP) is used to map MAC addresses to network IP addresses. the server checks its ARP table to see whether it has the MAC address of the router’s NIC. It then inspects the destination IP address in the packet and uses its routing . thereby ensuring that the packet is sent to the correct device. in which both parties can speak at the same time.03Harrison_ch02. It is for this reason that duplex settings aren’t usually a problem for Linux servers. its source/destination IP address stays the same.168. The MAC Address The media access control (MAC) address can be equated to the serial number of the NIC. MAC addresses therefore have significance only on the locally attached network. but the MAC addresses are reassigned by each router on the way using a process called ARP. the server issues an ARP request asking that router 192. let’s say with an IP address of 192. When a server needs to communicate with another server it takes the following steps: 1. is more like a walkie-talkie in which both parties have to wait until the other is finished before they can speak. the packet is sent and the ARP table is subsequently updated with the new MAC address. 2.1. leaving the MAC information behind. Most modern network cards can autonegotiate duplex with the device on the other end of the wire. If there is a valid router. the server sends the IP packet to its NIC and tells the NIC to encapsulate the packet in a frame destined for the MAC address of the router. Data transfer speeds will be low and error levels will be high if you have a device at one end of a cable set to full duplex and a device at the other end of the cable set to half duplex. 3. As each router in the path receives the packet. 5.1 respond with its MAC address so that the delivery can be made.qxd 2/25/05 9:55 AM Page 26 26 Introduction to Networking Chapter 2 A good analogy for full duplex communications is the telephone. on the other hand. Half duplex. it plucks the IP packet out of the Ethernet frame. You could very loosely view this as the server trying to find the Ethernet serial number of the next hop router on the local network.

9. ARP entries are not permanent and are erased after a fixed period of time depending on the operating system used. The Two Broad Types of Networking Equipment There are two main types of networking equipment: Data Communications Equipment (DCE).03Harrison_ch02. The ARP table is queried. The router then uses the “ARP-ing” process to get the MAC address of this next hop router. Data Communications Equipment DCE is also known as Data Circuit-Terminating Equipment and refers to such equipment as modems and other devices designed primarily to provide network access. . the ARP table contains only the MAC addresses of devices on the locally connected network. 8. The server will not send the data to its intended destination unless it has an entry in its ARP table for the next hop. which covers Linux network topics. It then reencapsulates the packet in an Ethernet frame with the new MAC address and sends the frame to the next hop router. the application needing to communicate will issue a timeout or time exceeded error. table to determine the IP address of the next router on the path to this destination. This relaying process continues until the packet reaches the target computer. If it doesn’t. a similar process occurs.qxd 2/25/05 9:55 AM Page 27 Networking Equipment Terminology 27 6. As can be expected. Data Terminal Equipment DTE devices were originally computer terminals located at remote offices or departments that were directly connected modems. The terminals would have no computing power and only functioned as a screen/keyboard combination for data processing. the packet is sent and the ARP table is subsequently updated with the new MAC address. Chapter 3. If the target server is on the same network as the source server. and Data Terminal Equipment (DTE). 7. which is intended to act as the primary communications path. an ARP request is made asking the target server for its MAC address. If no entry is available. When a reply is received. shows how to see your ARP table and the MAC addresses of your server’s NICs. which acts as the source or destination of the transmitted data. Nowadays most PCs have their COM and Ethernet ports configured as if they were going to be connected to a modem or other type of purely networking-oriented equipment.

Table 2. It is possible for multiple servers to speak at once. the crossover cable is often referred to as a null modem cable. The terminology can be different for cables used to connect serial ports together. When connecting a PC’s COM port (DTE) to a modem (DCE) the straight-through cable is frequently called a modem cable. Table 2. Hold the connectors side by side. confusion can arise when selecting a cable. A crossover cable has them mixed up. and other manufacturers have designed their equipment to flip automatically between the two types until it gets a good link. pointing in the same direction with the clips facing away from you. after a random time interval. or very little. until the message gets through correctly. .03Harrison_ch02. electronics inside and therefore do not regulate traffic. Hubs have no. try using the other type of cable. Hubs physically cross-connect all their ports with one another. Some manufacturers configure the Ethernet ports of their networking equipment to be either of the DTE or the DCE type. When connecting two PCs (DTE) together via their COM ports. A straight-through Ethernet cable is easy to identify. When this happens the servers try again. the same color for corresponding wires on each end. As you can see.qxd 2/25/05 9:55 AM Page 28 28 Introduction to Networking Chapter 2 Using Straight-Through/Crossover Cables to Connect DTEs and DCEs When a DCE is connected to a DTE. with all of them receiving garbled messages.3 Cabling Rules of Thumb Likely Cable Type Crossover Crossover Crossover Straight through Straight through Straight through Scenario PC to PC Hub to hub Switch to switch PC to modem PC to switch PC to hub Connectivity Using Hubs A hub is a device into which you can connect all devices on a network so that they can talk together. This terminology is generally used with Ethernet cables. The same would go for positions #2 through #8— that is.3 provides some good rules of thumb. DCEs connected to DCEs or DTEs connected to DTEs require crossover cables. which causes all traffic sent from a server to the hub to be blurted out to all other servers connected to that hub whether they are the intended recipient or not. you need a straight-through cable. If you fail to get a link light when connecting your Ethernet devices together. The color of the wire in position #1 on connector #1 should be the same as that of position #1 on connector #2.

more expensive switches can be configured to assign only certain ports to prespecified virtual LANs (VLANs) chosen by the network administrator. thereby eliminating the possibility of message garbling and providing a more efficient traffic flow. A switch is a much more reliable and predictable alternative. By connecting its NIC cards to multiple LANs. How Routers Interconnect LANs As stated before. which is blurted out to all the servers simultaneously. Routers are also capable of filtering traffic passing between the two LANs. It is possible to connect multiple switches and/or hubs in a chain formation to create a LAN with more ports. therefore providing additional security. A router still needs to be connected to each VLAN for internetwork communication. Unlike a hub. or tied to a wireless access point (WAP) that can communicate only with each other. Larger. a correctly configured router is capable of relaying traffic between networks. Switches regulate traffic. Servers use their IP address and subnet mask and the IP address of the remote server to determine whether they are both on the same network. and ones made for the home often cost only a few dollars more. Switches and hubs provide no access control between servers connected to the same LAN. the switch houses ports on multiple LANs. Note Hubs can add a lot of delays to your network because of the message garbling collisions and retransmissions. . If not. This is often called daisy chaining. they attempt to communicate with each other via routers that interconnect their LANs. Local Area Networks A local area network (LAN) is a grouping of ports on a hub. switches and hubs usually have only servers connected to them that have been configured as being part of the same network. The only exception is broadcast traffic. traffic sent from Server A to Server B will be received only by Server B.qxd 2/25/05 9:55 AM Page 29 Networking Equipment Terminology 29 It is for these reasons that Ethernet devices that plug into hubs should be set to half duplex. This is why network administrators group trusted servers having similar roles on the same LAN. In this case. Devices that plug into switches should be set to full duplex to take full advantage of the dedicated bandwidth coming from each switch port. Using Switches as a Faster Alternative to Hubs A switch is also a device into which you can connect all devices on a home network so that they can talk together.03Harrison_ch02. switch.

Chapter 3. a “route” refers to the path data takes to traverse from its source to its destination. it is possible to deny communication between two servers on different networks that intend to communicate on TCP port 80 and allow all other traffic between them. you must reserve an IP address for a router for each network and make sure that the router is directly connected to the LAN associated with that network. Therefore a home DSL router that provides IP Internet access to an Ethernet network is technically a gateway. we are referring to the IP address of the first hop needed to reach the desired destination network. Home Linux boxes frequently don’t run a dynamic routing protocol and therefore rely on static routes issued by the system administrator at the command line or in configuration files to determine the next hop to all desired networks. Usually when we speak about a route on a Linux box. Routers transfer traffic where both the protocols and communications medium are the same. Routers therefore direct and regulate traffic between separate networks. Each router along the way may also be referred to as a hop. shows how to add static routes to your Linux box and also how you can convert it into a simple router. The distinction can be important in complicated networking environments where newer technologies need to talk with older ones using incompatible communications protocols.03Harrison_ch02. much like a traffic policeman. It is assumed that this first hop will know how to automatically relay the packet.qxd 2/25/05 9:55 AM Page 30 30 Introduction to Networking Chapter 2 Routers can also be configured to deny communication between specific servers on different networks. routers most frequently provide connectivity to the Internet using network address translation or NAT. In home networks. Routers are designed to exchange routing information dynamically and can therefore intelligently redirect traffic to bypass failed network links. The terms are frequently used interchangeably. routers act as gateways to the wider world and it won’t be surprising to learn that routers are frequently referred to as “gateways.” Note The term gateway specifically refers to a device that routes traffic between dissimilar network protocols (IP to Appletalk) or access methods (Ethernet to DSL). How Simple Routing Works In the broader networking sense. which covers Linux network topics. especially if only one network protocol is being used. They can also filter traffic based on the TCP port section of each packet. . If you intend to route between networks. In other words. For example.

03Harrison_ch02. You could put a route on your SOHO servers that states: ☞ Go to network 10.0. For most home networks. Say for example you have two routers. not just by port and IP address as routers do. These VPNs are frequently used to connect branch offices to the corporate headquarters and also to allow sales representatives to get access to sensitive pricing information when traveling from town to town.qxd 2/25/05 9:55 AM Page 31 Networking Equipment Terminology 31 Default Gateways. Firewalls Help Provide a Secure Routing Environment Firewalls can be viewed as routers with more enhanced abilities to restrict traffic.0 via router R2.X via another NIC card. R1 and R2. A short list of capabilities includes the following: ☞ Throttling traffic to a server when too many unfulfilled connections are made to it ☞ Restricting traffic being sent to obviously bogus IP addresses ☞ Providing network address translation or NAT Routers are designed to make packets flow as quickly as possible with the minimum amount of inspection. R1 is connected to both your SOHO home network and the Internet. the Routers of Last Resort A default gateway is the router that is used when no alternative devices can be found to relay the traffic. . They are often called routers of last resort.0. your default gateway would be the router/ firewall connected to the Internet.0. Specifically. shows how to configure the default gateway on your Linux box. firewalls can detect malicious attempts to subvert the TCP/IP protocol. which covers Linux network topics. Firewalls can often create an encrypted data path between two private networks across the Internet providing secure communication with a greatly reduced chance of eavesdropping. Chapter 3.X.0. R2 is connected to SOHO home network and is capable of relaying data to other corporate networks with addresses starting with 10. R1 therefore would be considered your default gateway.X.0 255. Firewalls are used as close to the source or target of data communication as possible to try to ensure that the data hasn’t been subverted. ☞ Go to everything else via router R1.

” on RPM software installation. which gives a list of all the man pages that contain the word ssh: . “Troubleshooting Linux with Syslog. Knowledge of FTP is especially important and is a primary method of downloading software for Linux systems. The remote FTP server prompts you for a username. anonymous FTP. such as WSFTP and CuteFTP. If you want to get information on the ssh command. The second method.03Harrison_ch02. man -k ssh. at which point the user types anonymous or ftp with the password being any valid e-mail address. From the remote user’s perspective. you use the command man ssh: [root@bigboy tmp]# man ssh If you want to search all the man pages for a keyword. there are two types of FTP. From the command line you issue the man command followed by the Linux command or file about which you want information. The remote FTP server prompts you for a specific username and password to gain access to the data. You can also use FTP from the command line. is used primarily to allow any remote user to download files to their systems. and data are sent across the network unencrypted. Linux Help Linux help files are accessed using the man or manual pages. “Linux FTP Server Setup. “Secure Remote Logins and File Copying”) that is normally installed by default with Fedora. There are a number of commercially available GUI-based clients you can load on your PC to do this. passwords. These are active and passive FTP. use the man command with the -k switch—for example. The first is regular FTP. but they cover Linux help methods that you’ll use extensively and the File Transfer Protocol (FTP) package.” Remember that FTP isn’t very secure because usernames. From the systems administrator’s perspective. which is used primarily to allow specific users to download files to their systems. The File Transfer Protocol FTP is one of the most popular applications used to copy files between computers via a network connection. More secure forms—such as SFTP (Secure FTP) and SCP (Secure Copy)—are available as a part of the Secure Shell package (covered in Chapter 17.qxd 2/25/05 9:55 AM Page 32 32 Introduction to Networking Chapter 2 ADDITIONAL INTRODUCTORY TOPICS The last few topics of this chapter may not appear to be directly related to networking. as shown in Chapter 5. which enables you to download all the software you need to get your Linux server operational as quickly as possible. there are two more categories. which are covered in more detail in Chapter 15.

. Linux help can sometimes be cryptic.qxd 2/25/05 9:55 AM Page 33 Conclusion 33 [root@bigboy tmp]# man -k ..03Harrison_ch02. use its man page to get more details on what it does and the syntax it needs. CONCLUSION Congratulations! Now that you have an understanding of basic networking.. but with a little practice the man pages can become your friend. it’s time to read Chapter 3 to learn how to configure Linux networking... Feel free to return to this chapter whenever you need to refresh your memory on these foundation concepts. [root@bigboy tmp]# ssh - OpenSSH SSH client (remote login OpenSSH SSH client (remote login authentication agent gather ssh public keys OpenSSH SSH client configuration OpenSSH SSH daemon OpenSSH SSH daemon configuration program) program) files file This book is targeted at proficient Linux beginners and above. If you need more help on a command.. . .. ssh (1) ssh [slogin] (1) ssh-agent (1) ssh-keyscan (1) ssh_config (5) sshd (8) sshd_config (5) . .. so I’ll be using a wide variety of commands in this book without detailed explanations to help keep the flow brisk.

qxd 2/25/05 9:55 AM Page 34 .03Harrison_ch02.

04Harrison_ch03. Some of these activities are automatically covered during a Linux installation.qxd 2/25/05 9:56 AM Page 35 C H A P T E R 3 Linux Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ How to Configure Your NIC’s IP Address How to Activate/Shut Down Your NIC How to View Your Current Routing Table How to Change Your Default Gateway How to Configure Two Gateways How to Delete a Route Changing NIC Speed and Duplex How to Convert Your Linux Server Into a Router Configuring Your /etc/hosts File Conclusion Now that you have a firm grasp of many of the most commonly used networking concepts. so we’ll start this Linux networking chapter with a discussion on how to configure the IP address of your server. it is time to apply them to the configuration of your server. but you will often find yourself having to know how to modify these initial settings whenever you need to move your server to another network. add a new network interface card. or use an alternative means of connecting to the Internet. In Chapter 2. “Introduction to Networking. 35 .” we started with an explanation of TCP/IP.

255. You also might need to add a secondary NIC interface to your server to handle data backups.0 b) TX bytes:0 (0.1.168. Web site shopping cart applications frequently need an additional IP address dedicated to them. Last but not least.255 Mask:255.0 UP LOOPBACK RUNNING MTU:16436 Metric:1 RX packets:787 errors:0 dropped:0 overruns:0 frame:0 TX packets:787 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:0 RX bytes:82644 (80.255.0.1.168.04Harrison_ch03.1. you might just want to play around with the server to test your skills.4 Mb) TX bytes:43209032 (41.1.168. This section shows you how to do the most common server IP activities with the least amount of headaches.255.168. this device is called eth0.255 Mask:255. Determining Your IP Address Most modern PCs come with an Ethernet port.1 Mask:255.0 b) Interrupt:11 Base address:0x1820 lo Link encap:Local Loopback inet addr:127.255.7 Kb) TX bytes:82644 (80.0.0.100 Bcast:192.2 Mb) Interrupt:11 Memory:c887a000-c887b000 wlan0:0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.7 Kb) wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.qxd 2/25/05 9:56 AM Page 36 36 Linux Networking Chapter 3 HOW TO CONFIGURE YOUR NIC’S IP ADDRESS You need to know all the steps to configure IP addresses on a NIC card.0.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]# . When Linux is installed.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:47379 errors:0 dropped:0 overruns:0 frame:0 TX packets:107900 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:4676853 (4.99 Bcast:192. You can determine the IP address of this device with the ifconfig command: [root@bigboy tmp]# ifconfig -a eth0 Link encap:Ethernet HWaddr 00:08:C7:10:74:A8 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:0 (0.

which are then used to autoconfigure your NICs when Linux boots. You can look at the contents of the /proc/interrupts file to get a listing of all the interrupt IRQs used by your system.0 up The up at the end of the command activates the interface.1. eth0 has no IP address because this box is using wireless interface wlan0 as its main NIC Interface.04Harrison_ch03.255.255.1 netmask 255. One assumes the interface has a fixed IP address. You can place your IP address information in these files. eth1 uses ifcfg-eth1. Changing Your IP Address If you wanted. or PCI bus ID. wlan0 has an IP address of 192.qxd 2/25/05 9:56 AM Page 37 How to Configure Your NIC’s IP Address 37 In this example. and the other assumes it requires an IP address assignment using DHCP.255. and so on.168.1 for two samples of interface eth0. which is run at the end of every reboot. Interface eth0 has a file called ifcfg-eth0. respectively: [root@bigboy tmp]# cat /proc/interrupts CPU0 0: 2707402473 XT-PIC timer 1: 67 XT-PIC i8042 2: 0 XT-PIC cascade 5: 411342 XT-PIC eth1 8: 1 XT-PIC rtc 10: 1898752 XT-PIC eth0 11: 0 XT-PIC uhci_hcd 12: 58 XT-PIC i8042 14: 5075806 XT-PIC ide0 15: 506 XT-PIC ide1 NMI: 0 ERR: 43 [root@bigboy tmp]# If there are conflicts.100 and a subnet mask of 255. See Figure 3. Fedora Linux also makes life a little easier with interface configuration files located in the /etc/sysconfig/network-scripts directory.0. used by each card. On very rare occasions you may find that your NIC card doesn’t work because it shares both an interrupt and memory access address with some other device. To make this permanent each time you boot up you have to add this command in your /etc/rc. Devices eth0 and eth1 use interrupts 10 and 5. In the example below we can see that there are no conflicts with each IRQ from 0 to 15 having only a single entry.0 You can see that this command gives good information on the interrupts.255.0. you could give this ifconfig command: eth0 interface an IP address using the [root@bigboy tmp]# ifconfig eth0 10. you might need to refer to the manual for the offending device to try to determine ways to either use another interrupt or memory I/O location. .local file.

because the parameter has the value yes and not no. The ifdown and ifup commands can be used to do this: ONBOOT [root@bigboy network-scripts]# ifdown eth0 [root@bigboy network-scripts]# ifup eth0 Your server will have to have a default gateway for it to be able to communicate with the Internet.168. These are optional.168.qxd 2/25/05 2:33 PM Page 38 38 Linux Networking Chapter 3 Fixed IP Address [root@bigboy tmp]# cd /etc/sysconfig/network-scripts [root@bigboy network-scripts]# less ifcfg-eth0 DEVICE=eth0 IPADDR=192.1.100 NETMASK=255. .1.255. As you can see.0 [root@bigboy network-scripts]# Getting the IP Address Using DHCP [root@bigboy tmp]# cd /etc/sysconfig/network-scripts [root@bigboy network-scripts]# less ifcfg-eth0 DEVICE=eth0 BOOTPROTO=dhcp ONBOOT=yes [root@bigboy network-scripts]# Figure 3.255 NETWORK=192. This will be covered later in the chapter. which acts as an introduction to networking.168. You can read more about netmasks and DHCP in Chapter 2. After you change the values in the configuration files for the NIC you have to deactivate and activate it for the modifications to take effect.1. eth0 will be activated on booting.1 File formats for network-scripts.0 BOOTPROTO=static ONBOOT=yes # # The following settings are optional # BROADCAST=192.04Harrison_ch03. The default Red Hat/Fedora installation will include the broadcast and network options in the network-scripts file.255.

99 \ netmask 255. Multiple IP Addresses on a Single NIC In the previous section.04Harrison_ch03. The process for creating an IP alias is very similar to the steps outlined for the real interface in the previous section. 3.168. “Changing Your IP Address:” 1. Make sure your /etc/resolv. where X is the subinterface number of your choice. Create the virtual interface with the ifconfig command: [root@bigboy tmp]# ifconfig wlan0:0 192. Aliases have the name format parent-interface-name:X. .255.qxd 2/25/05 9:56 AM Page 39 How to Configure Your NIC’s IP Address 39 How DHCP Affects the DNS Server You Use Your DHCP server not only supplies the IP address your Linux box should use. Verify that no other IP aliases with the same name exist with the name you plan to use.1. Here is a sample configuration: DEVICE=wlan0:0 ONBOOT=yes BOOTPROTO=static IPADDR=192.255. IP aliasing is one of the most common ways of creating multiple IP addresses associated with a single NIC. Aliases can be shut down independently of other interfaces. After completing these four simple steps you should be able to ping the new IP alias from other servers on your network. Interface wlan0:0 is actually a child of interface wlan0. First ensure the parent real interface exists. “Determining Your IP Address. a virtual subinterface also known as an IP alias.99 NETMASK=255.” you may have noticed that there were two wireless interfaces: wlan0 and wlan0:0.168.1.0 The commands to activate and deactivate the alias interface would therefore be: [root@bigboy tmp]# ifup wlan0:0 [root@bigboy tmp]# ifdown wlan0:0 Note Shutting down the main interface also shuts down all its aliases. In this we want to create interface wlan0:0. but also the desired DNS servers.conf file has the servers configuration lines commented out to prevent any conflicts.0 up 4. 2. You should also create a /etc/sysconfig/network-scripts/ifcfg-wlan0:0 file so that the aliases will all be managed automatically with the ifup and ifdown commands.255.255.

04Harrison_ch03. you need to go through a number of steps to complete the connection. Your ISP will provide you with a PPP authentication over Ethernet (PPPoE) username and password that will allow your computer to log in transparently to the Internet each time it boots up. Here’s what you need to do: 1. remember that the PPPoE RPM’s filename usually starts with the word rp-pppoe followed by a version number like this: rp-pppoe3. Shut down your eth0 interface: [root@bigboy network-scripts]# ifdown eth0 [root@bigboy network-scripts]# . You plug your Ethernet interface into the DSL modem. Fedora Linux installs the rp-pppoe RPM software package required to support this.i386. The DHCP IP address assigned to your computer and/or Internet router will often not change for many days and you may be fooled into thinking it is static. subnet mask. Remember that you may also need to configure your DNS server correctly.qxd 2/25/05 9:56 AM Page 40 40 Linux Networking Chapter 3 IP Address Assignment for a Direct DSL Connection If you are using a DSL connection with fixed or static IP addresses. Note Unless you specifically request static IP addresses. DEVICE=eth0 ONBOOT=no 3. If you are using a DSL connection with a DHCP or dynamic IP address assignment. broadcast address. Edit your ifcfg-eth0 file to have no IP information and also to be deactivated on boot time. Make a backup copy of your ifcfg-eth0 file: [root@bigboy tmp]# [root@bigboy tmp]# cd /etc/sysconfig/network-scripts/ [root@bigboy network-scripts]# ls ifcfg-eth0 ifcfg-eth0 [root@bigboy network-scripts]# cp ifcfg-eth0 DISABLED. the process is different.rpm.” covers how to do this in detail. If you need a refresher. the configuration steps are the same as those outlined earlier. configure it with the IP address.5-8.ifcfg-eth0 2. After installing the RPM. Chapter 6. The PPPOE configuration will create a softwarebased virtual interface named ppp0 that will use the physical Internet interface eth0 for connectivity. “Installing RPM Software. your ISP will provide you with a DHCP-based connection. When searching for the file. Downloading and installing RPMs isn’t hard. and gateway information provided by your ISP and you should have connectivity when you restart your interface.

qxd 2/25/05 9:56 AM Page 41 How to Configure Your NIC’s IP Address 41 4. (default eth0): Do you want the link to come up on demand. Enter the demand value (default no): It will then prompt you for your DNS server information. Enter the DNS information here: The script will then prompt you for your ISP password: PASSWORD Please enter your Password: Please re-enter your Password: .. I will run some checks on your system to make sure the PPPoE client is installed properly. If you want your ISP to provide the IP address of its DNS server automatically. If you want the link to stay up permanently. Please refer to Chapter 18. If you’re running BIND on your server in a caching DNS mode then you may want to leave this option blank. LOGIN NAME Enter your Login Name (default root): bigboy-login@isp INTERFACE Enter the Ethernet interface connected to the ADSL modem For Solaris. If you just press enter. If your ISP claims that ‘the server will provide dynamic DNS addresses’. where ‘X’ is a number. DNS Please enter the IP address of your ISP’s primary DNS server. or stay up continuously? If you want it to come up on demand. it will be ethX.. We’ll use defaults wherever possible: Welcome to the ADSL client setup.) NOTE: Demand-activated links do not interact well with dynamic IP addresses. First. You may have some problems with demand-activated links. enter the word server. “Configuring DNS. enter ‘server’ (all lower-case) here. this is likely to be something like /dev/hme0. enter ‘no’ (two letters. For Linux. I will assume you know what you are doing and not modify your DNS setup. the interface to be used (eth0). lower-case. This step edits your /etc/resolv. and whether you want the connection to stay up indefinitely. Run the adsl-setup configuration script: [root@bigboy network-scripts]# adsl-setup It will prompt you for your ISP username.04Harrison_ch03. enter the idle time in seconds after which the link should be dropped.” for more information on BIND and DNS.conf file.

you’ll get a summary of the parameters you entered and the relevant configuration files will be updated to reflect your choices when you accept them: . however. Most people would say yes: Start this connection at boot time Do you want to start this connection at boot time? Please enter no or yes (default no):yes Just before exiting.NONE: This script will not set any firewall rules.04Harrison_ch03.” or the more comprehensive one found in the Appendix. “Linux Firewalls Using iptables. You are STRONGLY recommended to use some kind of firewall rules. the firewall rules will deny access to all standard servers like Web.) if you want to allow normal user to start or stop DSL connection (default yes): The rp-pppoe package has two sample iptables firewall scripts located in the /etc/ppp directory named firewall-standalone and firewall-masq. You are strongly encouraged to use a more sophisticated firewall setup.MASQUERADE: Appropriate for a machine acting as an Internet gateway for a LAN Choose a type of firewall (0-2): 0 You’ll then be asked whether you want the connection to be activated upon booting. the rules will block outgoing SSH connections which allocate a privileged source port. FIREWALLING Please choose the firewall rules to use. I’d recommend selecting none and using a variant of the basic script samples in Chapter 14.qxd 2/25/05 9:56 AM Page 42 42 Linux Networking Chapter 3 Then it will ask whether you want regular users (not superuser root) to be able to activate/deactivate the new ppp0 interface. lower-case. If you are running any servers on your machine. you must choose ‘NONE’ and set up firewalling yourself. Note that these rules are very basic. This may be required if non-root members of your family or home office need to get access to the Internet: USERCTRL Please enter ‘yes’ (three letters. They are very basic and don’t cover rules to make your Linux box a Web server. Otherwise. etc. ftp. 1 . these will provide basic security. You are responsible for ensuring the security of your machine. If you are using SSH. or mail server.STANDALONE: Appropriate for a basic stand-alone web-surfing workstation 2 . The firewall choices are: 0 . DNS server. e-mail.

This command defaults to show information for interface ppp0. Note This example recommends using the adsl-status command with the name of the PPPoE interface configuration file. After you have completed installing rp-pppoe you should be able to access the Internet over your DHCP DSL connection as expected. and therefore listing the ifcfg-ppp0 filename won’t be necessary in most home environments. The first is the ifcfg-ppp0 file with interface’s link layer connection parameters: [root@bigboy network-scripts]# more ifcfg-ppp0 USERCTL=yes BOOTPROTO=dialup NAME=DSLppp0 DEVICE=ppp0 TYPE=xDSL ONBOOT=yes PIDFILE=/var/run/pppoe-adsl. PPPOE_TIMEOUT=20 LCP_FAILURE=3 LCP_INTERVAL=80 . Some Important Files Created by adsl-setup The adsl-setup script creates three files that will be of interest to you.04Harrison_ch03.bak) At the very end it will tell you the commands to use to activate/deactivate your new ppp0 interface and to get a status of the interface’s condition: Congratulations.qxd 2/25/05 9:56 AM Page 43 How to Configure Your NIC’s IP Address 43 ** Summary of what you entered ** Ethernet Interface: eth0 User name: bigboy-login@isp Activate-on-demand: No DNS: Do not adjust Firewalling: NONE User Control: yes Accept these settings and adjust configuration files (y/n)? y Adjusting /etc/sysconfig/network-scripts/ifcfg-ppp0 Adjusting /etc/ppp/chap-secrets and /etc/ppp/pap-secrets (But first backing it up to /etc/ppp/chap-secrets.pid FIREWALL=NONE PING=.bak) (But first backing it up to /etc/ppp/pap-secrets. it should be all set up! Type ‘/sbin/ifup ppp0’ to bring up your xDSL link and ‘/sbin/ifdown ppp0’to bring it down. Type ‘/sbin/adsl-status /etc/sysconfig/network-scripts/ifcfg-ppp0’ to see the link status.

There are some good tips there on how to avoid problems with VPN clients. .POINTOPOINT. [root@bigboy tmp]# For further troubleshooting information you can visit the Web site of at Roaring Penguin (www.RUNNING..roaringpenguin..MULTICAST> mtu 1462 inet .04Harrison_ch03.. rp-ppoe .. adsl-status may be inaccurate. adsl-status: Link is attached to ppp0. In this case the package has been installed but the interface hasn’t been activated: [root@bigboy tmp]# adsl-status Note: You have enabled demand-connection. but ppp0 is down [root@bigboy tmp]# After activation.qxd 2/25/05 9:56 AM Page 44 44 Linux Networking Chapter 3 CLAMPMSS=1412 CONNECT_POLL=6 CONNECT_TIMEOUT=60 DEFROUTE=yes SYNCHRONOUS=no ETH=eth0 PROVIDER=DSLppp0 USER= bigboy-login@isp PEERDNS=no [root@bigboy network-scripts]# secrets The others are the duplicate /etc/ppp/pap-secrets and /etc/ppp/chapfiles with the username and password needed to login to your ISP: [root@bigboy network-scripts]# more /etc/ppp/pap-secrets # Secrets for authentication using PAP # client server secret IP addresses “bigboy-login@isp” * “password” [root@bigboy network-scripts]# Simple Troubleshooting You can run the adsl-status command to determine the condition of your connection. .com). the interface appears to work correctly: [root@bigboy tmp]# ifup ppp0 [root@bigboy tmp]# adsl-status adsl-status: Link is up and running on interface ppp0 ppp0: flags=8051<UP...

0.0 [root@bigboy tmp]# .0 are usually directly connected to the interface.0.qxd 2/25/05 9:56 AM Page 45 How to View Your Current Routing Table 45 IP Address Assignment for a Cable Modem Connection Cable modems use DHCP to get their IP addresses. the default and one to 255.0.255.04Harrison_ch03.0. Networks with a gateway of 0.0 0.0 0. ☞ In this example there are two gateways.0 0.255.0.255.255.0 127. You must have an ifcfg file in the /etc/sysconfig/network-scripts directory for these commands to work.0.168.0 Genmask 255. so a gateway address of 0.0.1.0.255.0.255 which is usually added on DHCP servers. HOW TO ACTIVATE/SHUT DOWN YOUR NIC The ifup and ifdown commands can be used respectively to activate and deactivate a NIC interface.255 255.255.255 0.0 255. Your default gateway is the route with a destination address of 0.0.0.0 Flags MSS Window irtt Iface UH 40 0 0 wlan0 U U UG 40 40 40 0 0 0 0 0 0 wlan0 lo wlan0 192.0 seems appropriate.0.0. so you can configure your server’s Ethernet interface accordingly.0.255. Server bigboy is a DHCP server in this case: [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Gateway 255.0.0.0 192.0.168.0.1 0.1. No gateway is needed to reach your own directly connected interface.255.0.0. Here is an example for interface eth0: [root@bigboy tmp]# ifdown eth0 [root@bigboy tmp]# ifup eth0 HOW TO VIEW YOUR CURRENT ROUTING TABLE The netstat -nr command will provide the contents of the routing table.0.

0. static IP addresses will need to have a manually configured default gateway.128 U UG UG U UG 172.69.67.69.0.0.69. there are multiple gateways handling traffic destined for different networks on different interfaces: [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Iface 172.0 172.16.16.64 172.0.16.0.96 172.0 172.1.16.67.255. you need to update your /etc/sysconfig/network file to reflect the change.255. This file is used to configure your default gateway each time Linux boots: NETWORKING=yes HOSTNAME=bigboy GATEWAY=192.1 .16.11.69.0.16.0 172.255.135 255.255.128 0.qxd 2/25/05 9:56 AM Page 46 46 Linux Networking Chapter 3 ☞ In this example.192 U 255.0 0.1 is connected to the same network as interface wlan0! When this is done.0 Gateway Genmask Flags MSS Window irtt 40 40 40 40 40 40 40 40 40 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 eth1 eth1 eth1 eth0 eth1 eth0 eth0 eth0 lo eth1 172.255.0 255.0.0 172.67.255.193 255.0.168.16.192 0.131 255.193 255.168.32 172.0.255.0.1.0.255.224 UG 255. most PCs would be using the standard Ethernet interface eth0: [root@bigboy tmp]# route add default gw 192.160.68.16.0 127.16.16.0 0.68.1.0.0 172. but NICs with configured.16. This example uses a newly installed wireless interface called wlan0.0 172. make sure that the router/firewall with IP address 192.69.255.1 wlan0 In this case.193 255.0.0.67.224 UG 172.255.0.16.255. DHCP servers automatically assign a default gateway to DHCP-configured NICs.0.04Harrison_ch03.67.135 255.168.0.193 0.0 [root@bigboy tmp]# HOW TO CHANGE YOUR DEFAULT GATEWAY Your server needs to have a single default gateway. This can be done with a simple command.224 UG 172.255.16.240.224 UG 172.255.16.

1.1: [root@bigboy tmp]# route add -host 10. In our example the reference to the 10. If one of the default gateways has no route to the intended destination.168.255.) Here is an example for a route to host 10.0.0. (The route command automatically knows the mask should be 255.255.0 \ gw 192.1 gw 192.254 wlan0 If you wanted to add a route to an individual server. This will cause packets to be sent to alternating gateways and could cause connectivity problems.168.0.0.04Harrison_ch03.254 wlan0 .168.0. Adding Routes from the Command Line The route add command can be used to add new routes to your server.255.0.0. respectively: [root@bigboy tmp]# route add -net 10.1. You may be most likely using interface eth0. Here’s a typical scenario: ☞ You have one router providing access to the Internet that you’d like to have as your default gateway (see the default gateway example earlier).d/rc.254 The Linux box used in this example uses interface wlan0 for its Internet connectivity.qxd 2/25/05 9:56 AM Page 47 How to Configure Two Gateways 47 Some people don’t bother with this step and just place the route add command in the script file /etc/rc. every other packet will become lost. Let’s assume that this router has an IP address of 192. It is possible to define default gateways in the NIC configuration file in the /etc/sysconfig/network-scripts directory.1.0.local which is run at the end of every reboot.0. It has the advantage of being universal to all versions of Linux and is well documented in the man pages. but you run the risk of inadvertently assigning more than one default gateway when you have more than one NIC. HOW TO CONFIGURE TWO GATEWAYS Some networks may have multiple router/firewalls providing connectivity.0.0 network has to be preceded with a -net switch. There are a number of ways to add this new route. Firewalls that are designed to block packets with irregular sequence numbers and unexpected origins could also obstruct your data flow.255.255. the -host switch would be used with no netmask value.0. please adjust your steps accordingly.0 to 10.255.0 netmask 255. and the subnet mask and gateway values also have to be preceded by the netmask and gw switches.0. ☞ You also have another router providing access to your corporate network using addresses in the range 10.

which then adds the route: [root@bigboy tmp]# netstat Kernel IP routing table Destination Gateway 192.1. There will also be no reported errors on the screen or any of the log files in the /var/log/ directory. In our routing example.0. which is always run at the end of the booting process.0./ifup-routes wlan0 [root@bigboy tmp]# netstat Kernel IP routing table Destination Gateway 192.0.168.0.0 0.0. Adding Routes with /etc/sysconfig/network-scripts/ Files In Fedora Linux.0.0.04Harrison_ch03.0 192.d/rc.0 255.0. permanent static routes are added on a per interface basis in files located in the /etc/sysconfig/network-scripts directory.0.254.0.0.0.0.0.0 0.0. Adding the wrong interface extension at the end will result in the routes not being added after the next reboot.255.254 Note The /etc/sysconfig/network-scripts/route-* filename is very important.0.0 0.0 0.254 0.0 169.1 [root@bigboy tmp]# -nr Genmask 255.0.0.0.local.0 0.0 0. The filename format is route-interface-name.1.0.0 with a subnet mask of 255.1.0.168.254. The format of the file is quite intuitive with the target network coming in the first column followed by the word via and then the gateway’s IP address. we would have to configure file /etc/sysconfig/network-scripts/route-wlan0 to look like this: # # File /etc/sysconfig/network-scripts/route-wlan0 # 10.0.0 0.0 Flags U U UG UG MSS 0 0 0 0 Window 0 0 0 0 irtt 0 0 0 0 Iface wlan0 wlan0 wlan0 wlan0 .0.0.0 169.0 255.0.0 (a mask with the first 8 bits set to 1) via the 192.1 [root@bigboy tmp]# -nr Genmask 255.168.1. In the next example we check the routing table to see no routes to the 10.0.254 gateway.1. You can test the new file by running the /etc/sysconfig/networkcommand with the interface name as the sole argument.168. so the filename for interface wlan0 would be route-wlan0.255.0 255.0 network and execute the ifup-routes command.0 10.168.0 192.168.0.1.0.1.0.0.0/8 via 192.0.qxd 2/25/05 9:56 AM Page 48 48 Linux Networking Chapter 3 A universal way of making this change persistent after a reboot would be to place this route add command in the file /etc/rc.0.255.255.168. to set up a route to network 10.0 192.0 Flags U U UG MSS 0 0 0 Window 0 0 0 irtt 0 0 0 Iface wlan0 wlan0 wlan0 scripts/ifup-routes [root@bigboy tmp]# .255.255.

but you have to refer to the manufacturer’s guides to set the speed and duplex to anything but auto-negotiate. Most Linux vendors have a list of compatible hardware on their Web sites.0.1. Don’t limit your troubleshooting of these types of errors to just failed negotiation. but this is usually accompanied by many collision-type errors being seen on the NIC.0. as seen in the next example. If you can’t find any of the desired models in your local computer store. when using the ifconfig -a command.0/8 via 192. so you need to be aware of it. but many older NICs support only mii-tool. Configuring a switch port to autonegotiate the speed and duplex often isn’t sufficient because there are frequently differences in the implementation of the protocol standard.0 \ gw 192. with unsupported NICs providing an Operation not supported message. Issuing the command without any arguments gives a brief status report. but only the fully compatible ones will provide you with errorfree.1. Typically.0.qxd 2/25/05 9:56 AM Page 49 Changing NIC Speed and Duplex 49 HOW TO DELETE A ROUTE Here’s how to delete the routes added in the previous section: [root@bigboy tmp]# route del -net 10. the server will not reinsert the route. NICs that are not compatible with mii-tool often will still work.168. It is destined to be deprecated and replaced by the newer ethtool command. Linux defaults to negotiating the speed and duplex of its NIC automatically with that of the switch to which it is attached. the problem could also be due to a bad NIC card. read this carefully before you start hooking up your machine to the network. Delete the line that reads 10. consistent throughput.0. a model in the same family or series should be sufficient. .254 CHANGING NIC SPEED AND DUPLEX There is no better Linux investment than the purchase of a fully Linux-compatible NIC card. NICs with failed negotiation will work.168. Most cards will work.0 netmask 255. or cabling.254 wlan0 The file /etc/sysconfig/network-scripts/route-wlan0 will also have to be updated so that when you reboot.0.04Harrison_ch03. and only marginal performance. Using mii-tool One of the original Linux tools for setting the speed and duplex of your NIC card was the mii-tool command.0. switch port.

model 33 basic mode: autonegotiation enabled basic status: autonegotiation complete. A list of supported modes is also provided at the top of the output: [root@bigboy tmp]# ethtool eth0 Settings for eth0: . link ok product info: vendor 00:10:18. The command provides the status of the interface you provide as its argument. or 10baseT-HD. full duplex. Here we see interface eth0 not doing autonegotiation and set to a speed of 100Mbps.04Harrison_ch03. negotiation was OK. capabilities: 100baseTx-FD 100baseTx-HD advertising: 100baseTx-FD 100baseTx-HD link partner: 100baseTx-FD 100baseTx-HD control [root@bigboy tmp]# rev 2 link ok 10baseT-FD 10baseT-HD 10baseT-FD 10baseT-HD 10baseT-FD 10baseT-HD flow- Setting Your NIC’s Speed Parameters with mii-tool You can set your NIC to force itself to a particular speed and duplex by using the -F switch with any of the following options: 100baseTx-FD. full duplex mode (FD): [root@bigboy tmp]# mii-tool -v eth1: negotiated 100baseTx-FD. “The Linux Boot Process.qxd 2/25/05 9:56 AM Page 50 50 Linux Networking Chapter 3 [root@bigboy tmp]# mii-tool SIOCGMIIPHY on ‘eth0’ failed: Operation not supported eth1: 100 Mbit. link ok [root@bigboy tmp]# By using the verbose mode -v switch you can get much more information. 100baseTx-HD. 10baseT-FD.local file to let it be run at the very end of the booting process or by creating your own startup script if you need it set earlier. with the NIC selecting 100Mbps.” Using ethtool The ethtool command is slated to be the replacement for mii-tool in the near future and tends to be supported by newer NIC cards. Creating your own startup scripts is covered in Chapter 7. Remember that you could lose all network connectivity to your server if you force your NIC to a particular speed/duplex that doesn’t match that of your switch: [root@bigboy tmp]# mii-tool -F 100baseTx-FD eth0 Unfortunately there is no way to set this on reboot permanently except by placing the command in the /etc/rc. In this case. half duplex.

In our next example. but you can get a quick guide by just entering the ethtool command alone.100 NETMASK=255. These settings can also be changed from the command line using the -s switch followed by the interface name and its desired configuration parameters. ethtool settings can be permanently set as part of the interface’s configuration script with the ETHTOOL_OPTS variable. [root@bigboy tmp]# ethtool -s eth1 speed 100 duplex full autoneg off [root@bigboy tmp]# The Linux man pages give more details on other ethtool options..04Harrison_ch03.1.. .255. which provides a quicker summary: [root@bigboy tmp]# ethtool . .qxd 2/25/05 9:56 AM Page 51 Changing NIC Speed and Duplex 51 Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Advertised auto-negotiation: No Speed: 100Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: internal Auto-negotiation: off Supports Wake-on: g Wake-on: g Current message level: 0x00000007 (7) Link detected: yes [root@bigboy tmp]# Setting Your NIC’s Speed Parameters with ethtool Unlike mii-tool. full duplex with no chance for autonegotiation on the next reboot: # # File: /etc/sysconfig/network-script/ifcfg-eth0 # DEVICE=eth0 IPADDR=192.255.0 BOOTPROTO=static ONBOOT=yes ETHTOOL_OPTS = “speed 100 duplex full autoneg off” You can test the application of these parameters by shutting down the interface and activating it again with the ifup and ifdown commands..168. the settings will be set to 100Mbps..

.. Remove the # from the line related to packet forwarding.ip_forward = 1 net. In simple terms packet forwarding enables packets to flow through the Linux box from one network to another. but when budgets are tight you might seriously want to consider modifying an existing Linux server to do the job.ipv4. at which time Linux will create a file in one of the subdirectories of the special RAM memory-based /proc filesystem. Here is how it’s done: [root@bigboy tmp] sysctl -p sysctl -p net.qxd 2/25/05 9:56 AM Page 52 52 Linux Networking Chapter 3 ethtool -s DEVNAME \ [ speed 10|100|1000 ] \ [ duplex half|full ] \ [ port tp|aui|bnc|mii|fibre ] \ .. Configuring IP Forwarding For your Linux server to become a router. but you need to understand how to activate routing through the firewall before it can become a functioning networking device.rp_filter = 1 . Before: # Disables packet forwarding net.ip_forward and can be found in the file /etc/sysctl.ipv4.ipv4. To activate the feature immediately you have to force Linux to read the /etc/sysctl.ipv4. Details on how to configure Linux firewall security are covered in Chapter 14.conf.ip_forward=0 After: # Enables packet forwarding net.default.conf file with the sysctl command using the -p switch. you have to enable packet forwarding. .conf. [root@bigboy tmp]# HOW TO CONVERT YOUR LINUX SERVER INTO A ROUTER Router/firewall appliances that provide basic Internet connectivity for a small office or home network are becoming more affordable every day. The Linux kernel configuration parameter to activate this is named net..ip_forward=1 This enables packet forwarding only when you reboot.04Harrison_ch03.ipv4.

These devices are typically inserted as part of a daisy chain connecting multiple network switches on the same LAN while protecting one section of a LAN from traffic originating on another section. If you need to enable proxy ARP on a Linux server the /proc filesystem comes into play again. In this scenario there is the possibility that either one will answer the local server’s ARP request for the MAC address of the remote server. “Miscellaneous Linux Topics. The server then sends an ARP request for the MAC address that matches the router’s IP address. There are some disadvantages with this. it sends out an ARP request to the network asking for the MAC address of the other device. If the same server needs to send a packet to another device on a remote network the process is different. It then sends the packet to the router using the router’s MAC address and a destination IP address of the remote server. If there is no suitable router on its network. Configuring Proxy ARP If a server needs to send a packet to another device on the same network. Some routers can be configured to answer these types of ARP requests for remote networks. The server first takes a look at its routing table to find out the IP address of the best router on its network that will be able to relay the packet to the destination. Each subdirectory . This feature is called proxy ARP.sysrq = 0 kernel.qxd 2/25/05 9:56 AM Page 53 How to Convert Your Linux Server into a Router 53 kernel. The firewall typically isn’t configured with an IP address on the LAN and appears to be an intelligent cable capable of selectively blocking packets. If one of the routers has an incorrect routing table entry for the remote network. In other words you can lose routing control.04Harrison_ch03. One of the most common problems occurs if two routers are on the network configured for proxy ARP. Note It is for this and other reasons that it is generally not a good idea to configure proxy ARP on a router. It is also good to always configure a default gateway on your server and use separate routing entries via other routers for all networks your default gateway may not know about. Proxy ARP is handled by files in the /proc/sys/net/ipv4/conf/ directory.” for more information on adjusting kernel parameters. Some types of bridging mode firewalls need to have proxy ARP enabled to operate properly. there is the risk that traffic to the remote server will occasionally get lost.core_uses_pid = 1 [root@bigboy tmp]# Please refer to Appendix I. This directory then has subdirectories corresponding to each functioning NIC card on your server. the server then sends out an ARP request for the MAC address of the remote server.

and use a centralized DNS server to handle most of the rest. Your server will typically check this file before referencing DNS. if the IP address for that host changes. The next example activates proxy ARP. You can use the /etc/sysctl. . but can become laborious if it has to be done companywide.conf # # Enables Proxy ARP on all interfaces net/ipv4/conf/all/proxy_arp = 1 # Enables Proxy ARP on interfaces eth1 and wlan0 net/ipv4/conf/eth1/proxy_arp = 1 net/ipv4/conf/wlan0/proxy_arp = 1 You can then activate these settings with the sysctl command: [root@bigboy tmp] sysctl -p CONFIGURING YOUR /ETC/HOSTS FILE The /etc/hosts file is just a list of IP addresses and their corresponding server names. For ease of management. server smallfry has an IP address of 192. or any other network-aware program by referring to it as smallfry.168.101) 56(84) bytes of data. proxy ARP on the interface is disabled.04Harrison_ch03.168.101 using ping. you also have to update the file. it is enabled.qxd 2/25/05 9:56 AM Page 54 54 Linux Networking Chapter 3 then has a file called proxy_arp. DNS won’t be queried at all.101.1. If the name is found with a corresponding IP address.1.168. Unfortunately. it is often easiest to limit entries in this file to just the loopback interface and also the server’s own hostname.1.101 smallfry In this example.conf file mentioned in Appendix II to activate or disable proxy ARP.1. and in such cases it may be easier to add a quick /etc/hosts file entry until the centralized change can be made: 192. Sometimes you may not be the one managing the DNS server. This may not be much of a concern for a single server.168. You can access 192. if the value is 1. Here is an example using the ping command to see whether smallfry is alive and well on the network: [root@bigboy tmp]# ping smallfry PING zero (192. If the value within this file is 0. first for all interfaces and then for interfaces eth0 and wlan0: # # File: /etc/sysctl. telnet.

Fedora inserts the hostname of the server between the 127.0.org localhost.1 needs to be the fully qualified domain name (FQDN) of the server— for example.1.168.qxd 2/25/05 9:56 AM Page 55 Configuring Your /etc/hosts File 55 64 bytes from smallfry (192.0.1.com: 127.1. this first entry after the 127. pipe 2 [root@bigboy tmp]# You can also add aliases to the end of the line.101): icmp_seq=1 ttl=64 time=0.092/0. 2 received.smallfry ping statistics --2 packets transmitted. time 2017ms rtt min/avg/max/mdev = 0.1 and the localhost entries like this: 127.0.localdomain localhost When the server is connected to the Internet.localdomain.0.168. By default. This is usually mapped to the name localhost. 0% packet loss.0.168.04Harrison_ch03.0.168.localdomain localhost Some programs such as Sendmail are very sensitive to this and if they detect what they feel is an incorrect FQDN they will default to using the name localhost. bigboy.1.1 bigboy. because the other server also feels it is localhost.1.074 ms.197 ms 64 bytes from smallfry (192.my-web-site. which enable you to refer to the server using other names.0.0. Here we have set it up so that smallfry can also be accessed using the names tiny and littleguy: 192.034/0.101 smallfry tiny littleguy You should never have an IP address more than once in this file because Linux will use only the values in the first entry it finds: 192.localdomain when communicating with another server on the network.1. .197/0.168.localdomain (the universal name used when a server refers to itself) and localhost (the shortened alias name).101 192.101 192.mysite.1 bigboy localhost.168. This can cause confusion.101): icmp_seq=0 ttl=64 time=0.101 smallfry tiny littleguy # (Wrong) # (Wrong) # (Wrong) The loopback Interface’s localhost Entry Usually the first entry in /etc/hosts defines the IP address of the server’s virtual loopback interface.047 ms --.

0.04Harrison_ch03. .qxd 2/25/05 9:56 AM Page 56 56 Linux Networking Chapter 3 Note You must always have a localhost and localhost. There always things that can go wrong that may be totally out of your control. Good systems administrators know the tools needed to be able to identify the probable causes of these types of problem. The skills you develop to identify and rectify these issues could prove to be invaluable to your company and career. configuring Linux networking is just a first step in providing Internet access to your server. The next two chapters show you how to test your network and Linux server applications confidently when things appear to go wrong. CONCLUSION As you can imagine.0.1 for Linux to work properly and securely.localdomain entry mapping to 127. which enables them to know the type of help they need to fix it.

qxd 2/25/05 9:56 AM Page 57 C H A P T E R 4 Simple Network Troubleshooting IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Doing Basic Cable and Link Tests Testing Your NIC How to See MAC Addresses Using ping to Test Network Connectivity Using telnet to Test Network Connectivity Testing Web Sites with the curl and wget Utilities The netstat Command The Linux iptables Firewall Using traceroute to Test Connectivity Using MTR to Detect Network Congestion Viewing Packet Flows with tcpdump Viewing Packet Flows with tethereal Basic DNS Troubleshooting Using nmap Determining the Source of an Attack Who Has Used My System? Conclusion You will eventually find yourself trying to fix a network related problem. which usually appears in one of two forms.05Harrison_ch04. These symptoms can be caused by: 57 . and the second is a complete lack of connectivity. The first is slow response times from the remote server.

☞ The switch or router to which the server is connected is powered down. “Introduction to Networking. In most cases a lack of link is due to the wrong cable type being used. More sophisticated models on the market will be able to tell you the approximate location of a cable break and whether an Ethernet cable is too long to be used. This indicates that the connection between your server and the switch/router is functioning correctly. ☞ The cables aren’t plugged in properly. . We discuss how to isolate these problems and more in the following sections.” and Chapter 19. “Configuring DNS. DOING BASIC CABLE AND LINK TESTS Your server won’t be able to communicate with any other device on your network unless the NIC’s link light is on.qxd 2/25/05 9:56 AM Page 58 58 Simple Network Troubleshooting Chapter 4 Sources of Network Slowness ☞ ☞ ☞ ☞ ☞ ☞ ☞ NIC duplex and speed incompatibilities Network congestion Poor routing Bad cabling Electrical interference An overloaded server at the remote end of the connection Misconfigured DNS (Covered in Chapter 18. Additional sources of disconnections are: ☞ Power failures ☞ The remote server or an application on the remote server being shut down.” there are two types of Ethernet cables: crossover and straight-through. As described in Chapter 2. “Dynamic DNS”) Sources of a Lack of Connectivity All sources of slowness can become so severe that connectivity is lost.05Harrison_ch04. Always make sure you are using the correct type. If you have an extensive network. investment in a battery-operated cable tester for basic connectivity testing is invaluable. Other sources of link failure include: ☞ The cables are bad.

119.10.255 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2295 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176.119.05Harrison_ch04. Interfaces will not appear if they are shut down: [root@bigboy tmp]# ifconfig Note Interfaces will appear if they are activated but have no link. whether they are functional or not. This can be seen in the next examples.4 Kb) Interrupt:10 Memory:c88b5000-c88b6000 Active Interface wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 inet addr:216. Shut Down Interface wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2287 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176. The following sections introduce a few commands that will be useful.6 Kb) Interrupt:10 Memory:c88b5000-c88b6000 .243 Bcast:216.qxd 2/25/05 9:56 AM Page 59 Testing Your NIC 59 TESTING YOUR NIC It is always a good practice in troubleshooting to be versed in monitoring the status of your NIC from the command line.7 Kb) TX bytes:166521 (162. Viewing Your Activated Interfaces The ifconfig command without any arguments gives you all the active interfaces on your system. Interfaces that are shut down by the systems administrator or are nonfunctional will not show an IP address line and the word UP will not show in the second line of the output. We’ll soon discuss how to determine the link status using commands.7 Kb) TX bytes:166377 (162.10. Viewing All Interfaces The ifconfig -a command provides all the network interfaces.

especially because mii-tool will be soon deprecated in Linux. Use it with the -v switch because it provides more information on the supported autonegotiation speeds of the NIC.qxd 2/25/05 9:56 AM Page 60 60 Simple Network Troubleshooting Chapter 4 Testing Link Status from the Command Line Both the mii-tool and ethtool commands provide reports on the link status and duplex settings for supported NICs. full duplex. The ethtool command provides much more information than mii-tool and should be your command of choice. link ok product info: Intel 82555 rev 4 basic mode: 100 Mbit. mii-tool gives a very brief report. In both of the following examples. full duplex. the NICs are operating at 100Mbps. full duplex basic status: link ok capabilities: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD advertising: 100baseTx-FD 100baseTx-HD 10baseT-FD 10baseT-HD flow-control link partner: 100baseTx-HD [root@bigboy tmp]# Link Status Output from ethtool [root@bigboy tmp]# ethtool eth0 Settings for eth0: Supported ports: [ TP MII ] Supported link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Supports auto-negotiation: Yes Advertised link modes: 10baseT/Half 10baseT/Full 100baseT/Half 100baseT/Full Advertised auto-negotiation: No Speed: 100Mb/s Duplex: Full Port: MII PHYAD: 1 Transceiver: internal Auto-negotiation: off Supports Wake-on: g Wake-on: g Current message level: 0x00000007 (7) Link detected: yes [root@bigboy tmp]# .05Harrison_ch04. and the link is ok. Link Status Output from mii-tool [root@bigboy tmp]# mii-tool -v eth0: 100 Mbit. When used without any switches. and this can be useful in troubleshooting speed and duplex issues.

Error rates in excess of 0. dropped packet.05Harrison_ch04.7 Kb) TX bytes:166377 (162. carrier. Ifconfig Error Output The ifconfig command also shows the number of overrun.qxd 2/25/05 9:56 AM Page 61 Testing Your NIC 61 Viewing NIC Errors Errors are a common symptom of slow connectivity due to poor configuration or excessive bandwidth utilization.4 Kb) Interrupt:10 Memory:c88b5000-c88b6000 ethtool Error Output The ethtool command can provide a much more detailed report when used with the -S switch: [root@probe-001 root]# ethtool -S eth0 NIC statistics: rx_packets: 1669993 tx_packets: 627631 rx_bytes: 361714034 tx_bytes: 88228145 rx_errors: 0 tx_errors: 0 rx_dropped: 0 tx_dropped: 0 multicast: 0 collisions: 0 rx_length_errors: 0 rx_over_errors: 0 rx_crc_errors: 0 rx_frame_errors: 0 rx_fifo_errors: 0 rx_missed_errors: 0 tx_aborted_errors: 0 tx_carrier_errors: 0 tx_fifo_errors: 0 tx_heartbeat_errors: 0 tx_window_errors: 0 tx_deferred: 0 tx_single_collisions: 0 tx_multi_collisions: 0 tx_flow_control_pause: 0 rx_flow_control_pause: 0 rx_flow_control_unsupported: 0 tx_tco_packets: 0 rx_tco_packets: 0 [root@probe-001 root]# . They should always be corrected whenever possible.5% can result in noticeable sluggishness. and frame errors. wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:D7 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:2924 errors:0 dropped:0 overruns:0 frame:0 TX packets:2287 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:180948 (176.

☞ Single Collisions: The Ethernet frame went through after only one collision. usually is an indication of electrical noise.qxd 2/25/05 9:56 AM Page 62 62 Simple Network Troubleshooting Chapter 4 Possible Causes of Ethernet Errors The following are possible causes of Ethernet errors: ☞ Collisions: The NIC card detects itself and another server on the LAN attempting data transmissions at the same time. The presence of CRC errors. and that the connectors are securely fastened. Taking a look at the ARP table of the server from which you are troubleshooting will help determine whether the remote server’s NIC is responding to any type of traffic from your Linux box. ☞ Carrier Errors: Errors are caused by the NIC losing its link connection to the hub or switch. but not many collisions. This is usually the result of collisions or a bad Ethernet device. ☞ Multiple Collisions: The NIC had to attempt multiple times before successfully sending the frame due to collisions. that the cabling is undamaged. Check for faulty cabling or faulty interfaces on the NIC and networking equipment. ☞ There might be bad network cabling. ☞ Frame Errors: An incorrect CRC and a noninteger number of bytes are received. This is usually a sign of excessive traffic. Collisions can be expected as a normal part of Ethernet operation and are typically below 0. HOW TO SEE MAC ADDRESSES There are times when you lose connectivity with another server that is directly connected to your local network.05Harrison_ch04.1% of all frames sent. Make sure that you are using the correct type of cable. ☞ A NIC might be disabled or the remote server might be shut down. ☞ Length Errors: The received frame length was less than or exceeded the Ethernet standard. . ☞ CRC Errors: Frames were sent but were corrupted in transit. ☞ FIFO and Overrun Errors: The number of times that the NIC was blocked from transferring data from the network to its memory buffers because of the speed limitations of the hardware. This is most frequently due to incompatible duplex settings. Lack of communication at this level may mean: ☞ Either server might be disconnected from the network. Higher error rates are likely to be caused by faulty NIC cards or poorly terminated cables.

168.168.101) at 00:06:25:09:6A:D7 [ether] on wlan0 [root@bigboy tmp]# ☞ Make sure the IP addresses listed in the ARP table match those of servers expected to be on your network.168.1. the server is running the correct software. .4 Mb) TX bytes:43209032 (41.168.255 Mask:255.1. but the desired communication doesn’t appear to be occurring to the client on the same network.1.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]# Here you can see the wlan0 interface has two IP addresses 192.168.168.1.2 Mb) Interrupt:11 Memory:c887a000-c887b000 wlan0:0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.168.255.1.05Harrison_ch04. Here we see we have some form of connectivity with the router at address 192.168. ☞ You should also check the ARP table of the remote server to see whether it is populated with acceptable values. Typically in this case.255.100 and 192.255 Mask:255.100 Bcast:192.255. your server may be plugged into the wrong switch or router port.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:47379 errors:0 dropped:0 overruns:0 frame:0 TX packets:107900 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:4676853 (4.1.1.168.qxd 2/25/05 9:56 AM Page 63 How to See MAC Addresses 63 ☞ The remote server might be running firewall software such as iptables or the Windows XP built-in firewall.99 tied to the NIC hardware MAC address of 00:06:25:09:6A:B5: ☞ The arp -a command will show you the MAC addresses in your server’s ARP table and all the other servers on the directly connected network. Here is a description of the commands you may use to determine ARP values: ☞ The ifconfig -a command shows you both the NIC’s MAC address and the associated IP addresses of the server that you are currently logged in to: [root@bigboy tmp]# ifconfig -a wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.99 Bcast:192.1. If they don’t.1: [root@bigboy tmp]# arp -a bigboypix (192.255. you can see the MAC address.1) at 00:09:E8:9C:FD:AB [ether] on wlan0 ? (192.1.

100: [root@smallfry tmp]# ping 192.101 ping statistics --4 packets transmitted.07 time=4. your traffic won’t be able to leave the LAN.1.1. One of the most common methods used to test connectivity across multiple networks is the ping command. A firewall or router along the network path is blocking ICMP traffic. 0% loss.168. This message is caused by your router or server knowing that the target IP address is part of a valid network. it becomes a very handy tool.1. Use traceroute to ensure you’re taking the correct path.948/7.46 time=4. but is getting no response from the target server.05Harrison_ch04. 64 bytes from 192. ping sends ICMP echo packets that request a corresponding ICMP echo-reply response from the device at the target address.31 ms ms ms ms --. There are a number of reasons for this.101: icmp_seq=2 64 bytes from 192.1. A lack of response could be due to: ☞ ☞ ☞ ☞ A server with that IP address doesn’t exist.1. Here is an example of a successful ping to the server bigboy at 192. Try pinging the gateways to make sure they are responsive. Because most servers will respond to a ping query. A classic symptom of bad routes on a server is the ability to ping servers only on your local network and nowhere else. . once a second.168. time 3026ms rtt min/avg/max/mdev = 3.168.101 PING 192.1. See the appendix for a full listing of them.168. If you can’t get to the first network hop.242 ms [root@smallfry tmp]# You may get a Destination Host Unreachable message.168.1. There are a variety of ICMP response codes that can help in further troubleshooting.1.101: icmp_seq=1 64 bytes from 192.192. The server has been configured not to respond to pings. ☞ You may have lost connectivity to the gateway router on the network.100 : 56(84) bytes ttl=128 ttl=128 ttl=128 ttl=128 time=3.101: icmp_seq=3 64 bytes from 192.950/4. it is always a good practice to force a response from it.95 time=7.168. until stopped with a Ctrl-C.168. You have incorrect routing.1.qxd 2/25/05 9:56 AM Page 64 64 Simple Network Troubleshooting Chapter 4 USING PING TO TEST NETWORK CONNECTIVITY Whether or not your troublesome server is connected to your local network.101: icmp_seq=4 192. ☞ Either the source or destination device has an incorrect IP address or subnet mask.101 (192. The Linux ping command sends continuous pings. Check the routes and subnet masks on both the local and remote servers and all routers in between.1.072/1.168.101) from of data. 4 received.168.168.

1.1.100 icmp_seq=6 Destination Host Unreachable --.168. +6 errors.102 22 When using telnet troubleshooting.1. There are two basic types: straight through and crossover. HTTPS uses port 443.1. which triggers the message. pipe 3 [root@smallfry tmp]# USING TELNET TO TEST NETWORK CONNECTIVITY An easy way to tell if a remote server is listening on a specific TCP port is to use the telnet command. The resulting message may be Destination Host Unreachable or Destination Network Unreachable: [root@smallfry tmp]# ping 192.168.1. HTTP uses TCP port 80.168.100 icmp_seq=4 Destination Host Unreachable From 192. By default. 0 received.168.100 : 56(84) bytes of data. ☞ Your NIC might not have the correct duplex settings.192.168. From 192.105 PING 192.100 icmp_seq=3 Destination Host Unreachable From 192.168.qxd 2/25/05 9:56 AM Page 65 Using telnet to Test Network Connectivity 65 If you are trying to ping a host on a directly connected network: ☞ The server might be down or disconnected for the network. you may verify this with the mii-tool command.100 icmp_seq=2 Destination Host Unreachable From 192. . but you can specify other TCP ports by typing them in after the target IP address.102 on the TCP port 22 reserved for SSH: [root@bigboy tmp]# telnet 192.1.168.1.105) from 192.168.05Harrison_ch04.1.1. telnet will try to connect on TCP port 23. ☞ In the case of a wireless network.100 icmp_seq=5 Destination Host Unreachable From 192.168. Here is an example of testing server 192. your SSID or encryption keys might be incorrect If you are trying to ping a host on a remote network: ☞ The network device doesn’t have a route in its routing table to the destination network and sends an ICMP reply type 3.168.1.168. here are some useful guidelines to follow that will help isolate the source of the problem: ☞ Test connectivity from the remote PC or server.1.168.1. ☞ You might have the incorrect type of cable connecting your Linux box to the network.105 ping statistics --8 packets transmitted.100 icmp_seq=1 Destination Host Unreachable From 192. 100% loss. time 7021ms.1.105 (192.168.

In some cases where the network is already protected by a firewall.05Harrison_ch04. You can use the /etc/init.102. This is often the cause of many connectivity problems and the firewall rules should be correctly updated. LINUX TELNET TROUBLESHOOTING telnet The following sections illustrate the use of Linux box..168. Successful Connection With Linux. Escape character is ‘^]’. but connectivity to desired TCP ports on the NIC interface may be blocked. all loopback connectivity could be allowed.168.102 on the SSH port (TCP 22): [root@bigboy tmp]# telnet 192. such as the Linux iptables software. .102. a successful telnet connection is always greeted by a Connected to message like this one when trying to test connectivity to server 192.qxd 2/25/05 9:56 AM Page 66 66 Simple Network Troubleshooting Chapter 4 ☞ Test connectivity on the server itself. troubleshooting from a Note Always remember that many Linux servers have the iptables firewall package installed by default. ☞ Test connectivity from another server on the same network as the target server. This helps eliminate the influence of any firewalls protecting the entire network from outside. Further discussion of the Linux iptables package is covered in a later section.1.1. SSH-1. If the server is running a firewall package. iptables may be safely turned off.. [root@ bigboy tmp]# To break out of the connection you have to press the Ctrl and ] keys simultaneously. Try making the connection to the loopback address as well as the NIC IP address.168.168. Connected to 192.102 22 Trying 192.1. not the usual Ctrl-C.1.d/iptables status command on the target server to determine whether iptables is running.99-OpenSSH_3.4p1 ^] telnet> quit Connection closed.

100: Connection refused [root@bigboy tmp]# telnet Timeout or Hanging The telnet command will abort the attempted connection after waiting a predetermined time for a response.. In some cases. Connection Refused Messages You will get a Connection refused message for one of the following reasons: ☞ The application you are trying to test hasn’t been started on the remote server. These symptoms can be caused by one of the following reasons: ☞ The remote server doesn’t exist on the destination network.1.. telnet won’t abort. causing it to time out instead of being quickly refused. .100. telnet: connect to address 192.10. [root@bigboy tmp]# telnet 216.12 22 Trying 216.12: Connection timed out [root@bigboy tmp]# TELNET TROUBLESHOOTING USING WINDOWS Sometimes you have to troubleshoot Linux servers from a Windows PC.12. Here is some sample output: [root@bigboy tmp]# telnet 192.10. This is called a timeout.05Harrison_ch04.168.100 22 Trying 192. This is usually caused by there being correct network connectivity but a poorly configured application.1.100. It could be turned off. but will just wait indefinitely. telnet: connect to address 216.1.10..168. The telnet commands are the same.qxd 2/25/05 9:56 AM Page 67 telnet Troubleshooting Using Windows 67 Note In many cases you can successfully connect on the remote server on the desired TCP port. Go to the command line and type the same telnet command as you would in Linux. yet the application doesn’t appear to work. ☞ A firewall could be blocking and not rejecting the connection attempt.168.100. ☞ There is a firewall blocking and rejecting the connection attempt. This is also known as hanging.100. but the results are different..

on port 256: Connect failed C:\> telnet Timeout or Hanging As explained previously.16. the session will appear to hang or time out.05Harrison_ch04.Could not open connection to the host.qxd 2/25/05 9:56 AM Page 68 68 Simple Network Troubleshooting Chapter 4 Screen Goes Blank—Successful Connection If there is connectivity.1. making troubleshooting difficult.. A much better combination of tools is to use telnet to test your site’s TCP port 80 response time in conjunction with data from the Linux curl and wget HTTP utilities. if there is no connectivity. your command prompt screen will go blank. Using curl The curl utility acts like a text-based Web browser in which you can select to see either the header or complete body of a Web page’s HTML code displayed on your screen.100. Using the Ctrl-C key sequence enables you to exit the telnet attempt. usually point not to a network issue.12 22 Connecting To 216. but slow curl and wget response times.102. but to slowness in the configuration of the Web server or any supporting application or database servers it may use to generate the Web page. C:\>telnet 216.10.. Rapid TCP response times. .100. Connect Failed Messages The Connect failed messages are the equivalent of the Linux Connection refused messages explained earlier and are caused by the same reasons: C:\>telnet 172.16.12..1. Many useful HTTP error codes are often not displayed by browsers..10. This is usually caused by the target server being turned off or by a firewall blocking the connection.102 256 Connecting To 172. TESTING WEB SITES WITH THE CURL AND WGET UTILITIES Testing a Web site’s performance using a Web browser alone is sometimes insufficient to get a good idea of the source of slow Web server performance.

115. Using wget You can use wget to recursively download a Web site’s pages.0.71.http://www.. awaiting response. 200 OK Length: unspecified [text/html] [ <=> 279.http://www.com/ => `index.34]:80. Connecting to www. you view not only the HTML content of the Web site’s index page in your local directory. This can be very helpful in providing a simple way to obtain snapshots of your server’s performance: [root@zippy tmp]# wget -N www. awaiting response. 200 OK Length: unspecified [text/html] Last-modified header missing -. connected. but also the download speed.51 (Fedora) Accept-Ranges: bytes Vary: Accept-Encoding.linuxhomenetworking.150 23:07:22 (279.linuxhomenetworking..34]:80. 19 Oct 2004 05:11:22 GMT Server: Apache/2.html’ Resolving www.qxd 2/25/05 9:56 AM Page 69 Testing Web Sites with the curl and wget Utilities 69 A good start is to use the curl command with the -I flag to view just the Web page’s header and HTTP status code.com/ => `index... By not using recursion and activating the time stamping feature (the -N switch). Either method can provide a good idea of your server’s performance: [root@ bigboy tmp]# curl -I www.User-Agent Connection: close Content-Type: text/html.115.71.. By not using the -I flag you will see all the Web page’s HTML code displayed on the screen.36K/s ] 122. Please refer to Chapter 20. HTTP request sent.05Harrison_ch04.linuxhomenetworking.com --23:07:22-. charset=UTF-8 [root@bigboy tmp]# In this case the Web server appears to be working correctly because it returns a 200 OK code. connected. HTTP request sent. “The Apache Web Server.linuxhomenetworking.time-stamps turned off. file size.com HTTP/1. done..linuxhomenetworking.linuxhomenetworking.36 KB/s) .html’ saved [122150] [root@zippy tmp]# ..com[65.`index.1 200 OK Date: Tue..” for a more complete listing of possibilities.com..html’ Connecting to www.com[65. to a local directory.. --23:07:22-. and precise start and stop times for the download. including the entire directory structure of the Web site.linuxhomenetworking.

193 0. It is always good to ensure that your routes are correct and that you can ping all the gateways in your routing table.240. can then be used to verify that your routing table is correct by displaying the path a packet takes to get to a remote destination. It is therefore a good idea to focus on these types of short-lived connections.0 172.0.0:* tcp 0 0 :::80 :::* .0.67..0 255.160. The traceroute command.255. .64 172.193 255.69.0 0.05Harrison_ch04. which I’ll discuss later.16.68.0.69.224 172.224 172.68.128 0.16. then your routing table needs to be examined more carefully.69.0 172.qxd 2/25/05 9:56 AM Page 70 70 Simple Network Troubleshooting Chapter 4 THE NETSTAT COMMAND Like curl and wget.255.128 172.135 255. which in this case shows 14 connections: [root@bigboy tmp]# netstat -an | grep tcp | egrep -i \ ‘established|time_wait’ | wc -l 14 [root@bigboy tmp]# The netstat -nr command can also be used to view your routing table.0 255. netstat can be very useful in helping determine the source of problems.16.224 172. Using netstat with the -an option lists all the TCP ports on which your Linux server is listening.32 172.0 172.16.131 255.0.. You can determine the number of established and time_wait TCP connections on your server by using the netstat command filtered by the grep and egrep commands.0 0.16.193 255.0. [root@bigboy tmp]# netstat -nr Kernel IP routing table Destination Gateway Genmask Iface 172.0.0. with the number of matches being counted by the wc command.16.16.69.224 172.67.96 172.16.0.0 255.0.192 0.255.11. [root@bigboy tmp]# State LISTEN LISTEN Most TCP connections create permanent connections.255.255.255.135 255.255.0.0.255.0.0.16. HTTP is different because the connections are shut down on their own after a predefined inactive timeout or time_wait period on the Web server.67.255. If the first hop is incorrect. ..16..255.16.0 172.69.16.255.0.192 172.0.0.0. including all the active network connections to and from your server.67.0.16..0 127.1:25 0.255.0.0.67.. This can be very helpful in determining whether slowness is due to high traffic volumes: [root@bigboy tmp]# netstat -an Active Internet connections (servers and established) Proto Recv-Q Send-Q Local Address Foreign Address tcp 0 0 127.16.0 [root@bigboy tmp]# Flags MSS Window irtt UG UG UG UG U U UG UG U UG 40 40 40 40 40 40 40 40 40 40 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 eth1 eth1 eth1 eth0 eth1 eth0 eth0 eth0 lo eth1 .0.0 172.255.193 255.

but also sends an ICMP time exceeded message back to the source.” USING TRACEROUTE TO TEST CONNECTIVITY Another tool for network troubleshooting is the traceroute command. The traceroute program records the IP address of the router that sent the message and knows that .d/iptables script with the stop qualifier: [root@bigboy tmp]# service iptables stop Flushing firewall rules: [ OK ] Setting chains to policy ACCEPT: filter [ Unloading iptables modules: [ OK ] [root@bigboy tmp]# OK ] How to Configure iptables Rules Stopping iptables may not be a good permanent solution. [root@zero root]# How to Stop iptables If your Linux box is already protected by a firewall. Determining Whether iptables Is Running You can easily test whether iptables is running by using the /etc/init. The traceroute command works by sending a UDP packet destined to the target with a TTL of 0. Here is some sample output: [root@zero root]# service iptables status Firewall is stopped.05Harrison_ch04.d/ iptables script with the status qualifier.qxd 2/25/05 2:34 PM Page 71 Using traceroute to Test Connectivity 71 THE LINUX IPTABLES FIREWALL An unexpected source of server connectivity issues for brand new servers is frequently the iptables firewall. It gives a listing of all the router hops between your server and the target server. This helps you verify that routing over the networks in between is correct. You can read more about configuring iptables in Chapter 14. you may want to temporarily disable iptables using the same /etc/init. “Linux Firewalls Using iptables. especially if your network isn’t protected by a firewall. This is installed by default under Fedora and Red Hat and usually allows only a limited range of traffic. The first router on the route recognizes that the TTL has already been exceeded and discards or drops the packet. If it isn’t running you’ll get a very short listing of the firewall rules.

232.another-isp-provider.17) 12. This continues around and around until the final destination is reached. The first hop sees nothing wrong with the packet.net (64.sntc01.618 ms 15.67) 13.260 ms 15.203.05Harrison_ch04.306 ms 26.408 ms 14.111 ms 2 dist3-vlan50.another-isp-provider.203.net (63.my-isp-provider.064 ms 13.sntc01.482 ms 26.443 ms 16. Note In Linux the traceroute command is traceroute.net (63.sntc01. drops the packet and replies with an ICMP time exceeded message.854 ms 13. 30 hops max.124. with a TTL of 1.49) 14.663 ms 5 bb1-p14-0.158 (144.232.35.785 ms 17.164 ms 6 sl-gw11-sj-3-0.20. The traceroute program tries again.887 ms 13. If a device responds.dsl. Notice that all the hop times are under 50 milliseconds (ms) which is acceptable: [root@bigboy tmp]# traceroute -I 144.110) 14. traceroute now knows the IP address of the second router.279 ms 15.20.185 ms 15.44. decrements the TTL to 0 as expected.snfc21.158) 27. 38 byte packets 1 adsl-67-120-221-110.158).124.161.232. and forwards the packet to the second hop on the path.487 ms [root@bigboy tmp]# Possible traceroute Messages There are a number of possible message codes listed in Table 4.net (144.146 ms 3 bb1-g1-0.246) 16.35.120. Router 2 sees the TTL of 0. Note You receive traceroute responses only from functioning devices.232.qxd 2/25/05 9:56 AM Page 72 72 Simple Network Troubleshooting Chapter 4 that is the first hop on the path to the final destination.745 ms 4 bb2-p11-0.232.snfc21. it is less likely to be the source of your problems.189 ms 7 sl-bb25-sj-6-1.035 ms 13.another-isp-provider. Sample traceroute Output Here is a sample output for a query to 144.3.221. traceroute can give.net (67.897 ms 15. In Windows it is tracert.423 ms 8 sl-bb23-ana-6-0.20.857 ms 15.net (144.133) 16.net (144.my-isp-provider.20.my-isp-provider.53) 15.161.1.net (64.20.232.018 ms 12.158.158 traceroute to 144.my-isp-provider.228. these are .my-isp-provider.

or protocol unreachable. network. !N.179 ms 4 bb2-p11-0.971 ms 14. Could be caused by the following: ☞ A router on the path not sending back the ICMP time exceeded messages ☞ A router or firewall in the path blocking the ICMP time exceeded messages ☞ The target IP address not responding Host.my-isp-provider. Source routing attempts to force traceroute to use a certain path.203. Source route failed.net (67. In this case the * * * status messages disappear: .842 ms 15.232.212 ms 14.net (63.304 ms 14. Communication administratively prohibited. or !P !X or !A !S traceroute Time Exceeded False Alarms If there is no response within a 5-second timeout interval.another-isp-provider.627 ms 3 bb1-g1-0.124. 38 byte packets 1 adsl-67-120-221-110.133) 16.232.my-isp-provider. !H.120 ms 2 dist3-vlan50.232.20.158).dsl.510 ms 15.snfc21.510 ms 17.991 ms 15.sntc01.net (144.net (144.44.sntc01.another-isp-provider.274 ms 15.158 (144.246) 13. A router Access Control List (ACL) or firewall is in the way.909 ms 6 sl-gw11-sj-3-0.53) 16.203.860 ms 13.20.net (64.17) 15.35.469 ms 18.728 ms 5 bb1-p14-0. Failure may be due to a router security setting.019 ms 16.228. as seen in the following example: [root@bigboy tmp]# traceroute 144.158 traceroute to 144.my-isp-provider. an asterisk (*) is printed for that probe.000 ms 14.116 ms 7 sl-bb25-sj-6-1.120.35.110) 14.161.snfc21.qxd 2/25/05 9:56 AM Page 73 Using traceroute to Test Connectivity 73 Table 4.521 ms 12.my-isp-provider. Using traceroute with an -I flag forces traceroute to use ICMP packets that may go through.926 ms 8 * * * 9 * * * [root@bigboy tmp]# Some devices will prevent traceroute packets directed at their interfaces.49) 16.20.124.net (64.232. but will allow ICMP packets. 30 hops max.221.161.05Harrison_ch04.my-isp-provider.3.sntc01.133 ms 15.1 traceroute Return Code Symbols Traceroute Symbol *** Description Expected 5-second response time exceeded.67) 12.net (63.

232.net (63.net (67.118. all points past the problematic link would show high latency.05Harrison_ch04.my-isp-provider. latency. Many Internet routing devices give very low priority to traffic related to traceroute in favor of revenue-generating traffic.253 192.20.146 ms 3 bb1-g1-0.net (144.232.164 ms 6 sl-gw11-sj-3-0.sntc01.203.110 192.my-isp-provider.net (144.net (64.227 Trace complete.482 ms 26.246) 16.net (63.net (144.120.49) 14.618 ms 15.064 ms 13.174.34 192.6 80.65 64.53) 15.443 ms 16.40. C:\> This indicates only that the devices on hops 6 and 7 were slow to respond with ICMP TTL exceeded messages.035 ms 13.018 ms 12.200.another-isp-provider.200.158).20.17) 12.another-isp-provider.44.134.250.200.161.110) 14.279 ms 15.158 (144.111 ms 2 dist3-vlan50.854 ms 13.35.12 80.229 64.96.35.185 ms 15.857 ms 15.189 ms 7 sl-bb25-sj-6-1.887 ms 13.158) 27.228. .40.another-isp-provider.150. 38 byte packets 1 adsl-67-120-221-110.663 ms 5 bb1-p14-0.my-isp-provider.net (64. or packet loss.21.203.124.sntc01.dsl.qxd 2/25/05 9:56 AM Page 74 74 Simple Network Troubleshooting Chapter 4 [root@bigboy tmp]# traceroute -I 144.40.174.158 traceroute to 144.snfc21.423 ms 8 sl-bb23-ana-6-0.897 ms 15.20.227 1 2 3 4 5 6 7 8 9 10 11 1 43 15 26 38 239 254 24 91 17 30 ms ms ms ms ms ms ms ms ms ms ms 2 15 16 13 12 255 252 20 89 20 16 ms ms ms ms ms ms ms ms ms ms ms 1 44 8 16 14 253 252 20 60 20 23 ms ms ms ms ms ms ms ms ms ms ms 66.118.31.118. 30 hops max.sntc01.149.255.133) 16.67) 13.487 ms [root@bigboy tmp]# traceroute Internet Slowness False Alarm The following traceroute gives the impression that a Web site at 80.20.221. If any of those conditions existed.193 64.232.161.306 ms 26.227 might be slow because there is congestion along the way at hops 6 and 7 where the response time is over 200ms: C:\>tracert 80.my-isp-provider.408 ms 14.200.785 ms 17.150.40.260 ms 15.200.151.3.745 ms 4 bb2-p11-0.47.my-isp-provider. but not an indication of congestion.97 172.snfc21.232.168.232.124.14 64.

☞ The server is shut down or disconnected from the network. but because that was the first hop at which the return packet traveled back to the source via the congested route. latency showed up at hop 7 with the return path being very different. ☞ The server is running some type of firewall software that blocks traceroute. A high traceroute time equates to the round-trip time for both the initial traceroute query to each hop and the response of each hop. a problem seemed to appear at the 10th hop with OtherNet. The problem is not with the router. Here is an example of one such case.68.100.168.40.05Harrison_ch04.18 Tracing route to 80.25.100.1. When a user at IP address 40. the real traffic congestion was occurring where FastNet handed off traffic to SlowNet in the second trace.1 80.175.14. Always Get a Bidirectional traceroute It is always best to get a traceroute from the source IP to the target IP and also from the target IP to the source IP.175. Remember. The latency appeared to be caused at hop 10 on the first trace not because that hop was slow. or it has an incorrectly configured NIC.25.19 81.16.65.97. C:\>tracert 80. However.18 over a maximum of 30 hops 1 33 2 33 3 33 4 47 5 29 6 * 7 ^C C:\> ms ms ms ms ms 49 49 32 32 28 * ms ms ms ms ms 28 28 32 31 32 * ms ms ms ms ms 192.200.32.252 80. In this case. using disguised IP addresses and provider names.40.114 Request timed out. Possible causes of this problem include the following: ☞ The server has a bad default gateway.96.16.200 did a traceroute to 40. you will only receive traceroute responses from functioning devices.qxd 2/25/05 9:56 AM Page 75 Using traceroute to Test Connectivity 75 traceroute Dies at the Router Just Before the Server In this case the last device to respond to the traceroute just happens to be the router that acts as the default gateway of the server. Remember. when a user at 64. traceroute gives the packet round-trip time: .40. This is because the packet’s return path from the target is sometimes not the same as the path taken to get there.1 65.25. There was once a routing issue between telecommunications carriers FastNet and SlowNet.106.106.32 did a traceroute to 64.40. but with the server.

webhostinc2.99] ms [64.qxd 2/25/05 9:56 AM Page 76 76 Simple Network Troubleshooting Chapter 4 Trace route to 40.net [4.cointcorp.253] ms [64.61] 13 1031 ms 989 ms 978 [4.25.38] 8 30 ms 30 ms 30 p15-0.225] 9 30 ms 30 ms 30 p1-0.200] .2] [40.25.30] ms p4-3.another-isp-provider.another-isp-provider.net sjo-core-01.24.25.p4-2.25.0.inet.net [4.another-isp-provider.226] 15 1027 ms 1025 ms 1023 [4.150] 10 1252 ms 1212 ms 1202 h0.9.net ms gw234.0.128.inet.dnvtco1-cr3.26.9.snjpca1-br1.6.16.5.24.dnvtco1-br1.171.25.net [64.fastnet.sfj.24.128.97] ms [205.another-isp-provider.96.16.11.77] 6 2 ms 1 ms 1 [216.22.11.26] 17 1030 ms 1020 ms 1045 [4.162] 13 1102 ms 1091 ms 1092 [40.106.snjpca1-br2.another-isp-provider.106.161] [40.25.net [207.130] 14 1031 ms 1017 ms 1017 [4.2.webh02-1.dnvtco1-br1.snjpca1-cr4.32 from 64.19.05Harrison_ch04.another-isp-provider.29] ms svl-core-03.another-isp-provider.ge0-1-net2.169] 4 0 ms 0 ms 0 [64.16.3] [40.253] 3 0 ms 0 ms 0 border-from-40-tesser.net ms p9-0.32 1 1 ms 1 ms 1 2 1 ms 1 ms 1 3 2 ms 1 ms 1 4 1 ms 1 ms 1 5 2 ms 2 ms 1 [216.38] 11 1252 ms 1212 ms 1192 [40.236.net [4.dnvtco1-br2.174.171.223.another-isp-provider.175.200 1 0 ms 0 ms 0 [64.net [205.net ms p6-0.96.24.106.another-isp-provider.net ms p1-0.49.126] 5 0 ms 0 ms 0 p3-0.37] 16 1045 ms 1037 ms 1050 [4.my-isp-provider.7.fastnet.dnvtco1-cr3.52.snjpca1-br2.net [4.net ms core1.11] 12 1262 ms 1212 ms 1192 [40.net ms sjo-edge-03.53] 6 0 ms 0 ms 0 p2-1.46.0.paix-bi1.171.175.175.another-isp-provider.25] 7 0 ms 0 ms 0 p15-0.net [4.96.200] 2 0 ms 0 ms 0 [64.96.slownet.54] 18 1038 ms 1031 ms 1045 19 1050 ms 1094 ms 1034 20 1050 ms 1094 ms 1034 ms ms ms ms ms [40.24.16.inet.16.net ms p15-0.16.another-isp-provider.24.33] 1009 ms 1008 ms 971 ms 9 985 ms 947 ms 983 10 1028 ms 1010 ms 953 11 989 ms 988 ms 985 [4.24.96.175.25.175.net ms p3-0.11.175.13] 12 1002 ms 1001 ms 973 [4.16.slownet.dnvtco1-br2.slownet.another-isp-provider.11.my-isp-provider.net [4.25.net [205.32] Trace route to 64.sfj.25.52.144.another-isp-provider.16.16.net ms p0-0.26.65] border8.200 from 40.6.25.65] 7 993 ms 961 ms 999 8 [208.205.106.24.

a ping to 186.my-isp-provider.64.64.94 and 186.9.net [186.17.64.1 rtr-2.17.9.9.my-isp-provider.net rtr-1.05Harrison_ch04.40. Average = 0ms G:\>tracert 186.64. ms ms ms ms ms ms ms ms ms <10 60 70 60 70 61 70 60 70 ms ms ms ms ms ms ms ms ms 186.. .net rtr-2.94: 186.153] over a maximum of 30 hops: 1 <10 ms <10 2 60 ms 70 3 70 ms 71 4 60 ms 70 5 70 ms 70 6 60 ms 70 7 70 ms 70 8 60 ms 70 9 70 ms 70 .40. transit.40..93] [186.. The routing loop was confirmed by the traceroute.40.40. .153 Tracing route to lostserver.94: TTL TTL TTL TTL expired expired expired expired in in in in transit.64.33. transit.qxd 2/25/05 9:56 AM Page 77 Using traceroute to Test Connectivity 77 ping and traceroute Troubleshooting Example In this example.94] [186.153: Packets: Sent = 4.153 gave a TTL timeout message.93] This problem was solved by resetting the routing process on both routers.64. Received = 4.17.217.my-isp-provider.17. traceroute Web Sites Many ISPs will provide their subscribers with the facility to do a traceroute from purpose-built servers called looking glasses.64.my-isp-provider.40. Ping statistics for 186.94] [186.93] [186. The problem was initially triggered by an unstable network link that caused frequent routing recalculations.64.40.93: G:\>ping 186.. Doing a .94: 186. Trace complete.17. A simple Web search for the phrase Internet looking glass will provide a long list of alternatives.net rtr-2.9. The constant activity eventually corrupted the routing tables of one of the routers.my-isp-provider. Ping TTLs will usually timeout only if there is a routing loop in which the packet bounces between two routers on the way to the target.my-isp-provider.my-isp-provider.40.94] [186. Approximate round trip times in milli-seconds: Minimum = 0ms.64. in which the packet was proven to be bouncing between routers at 186.net rtr-2. Each bounce causes the TTL to decrease by a count of 1 until the TTL reaches 0.153 Pinging 186.64.64.17.net [186.40.net rtr-1.40.94] [186..40.9..my-isp-provider. Maximum = 0ms.93] [186.40.153 with 32 bytes of data: Reply Reply Reply Reply from from from from 186.9.64. at which point you get the timeout.64.64. transit.net rtr-1.40.94: 186.40. Lost = 0 (0% loss).net rtr-1.my-isp-provider.

which provides a list of looking glasses sorted by country. ☞ Do a traceroute from this router to your intended target server. It’s usually good to check the routing table and/or other status of this next hop device. ☞ If this works: Routing to the target server is OK. The traceroute will probably fail at the bad router on the return path. Possible Reasons for a Failed traceroute A traceroute can fail to reach its intended destination for a number of reasons including the following: ☞ packets are being blocked or rejected by a router in the path. The target server doesn’t exist on the network. The last visible hop is the last hop in which the packets return correctly. Note If there is nothing blocking your traceroute traffic.org. ☞ Look at the routing table to determine what the next hop is to your intended traceroute target. It could be disconnected or turned off. The packets don’t have a proper return path to your server. The router immediately after the last visible one is usually the culprit. .) You may have a typographical error in the IP address of the target server. ☞ If it doesn’t work: Test the routing table and/or other status of all the hops between it and your intended target. Do a traceroute back to your source server.qxd 2/25/05 9:56 AM Page 78 78 Simple Network Troubleshooting Chapter 4 traceroute from a variety of locations can help identify whether the problem is with the ISP of your Web server or the ISP used at home/work to provide you with Internet access.) The network on which you expect the target host to reside doesn’t exist in the routing table of one of the routers in the path. the last visible router of an incomplete trace is either the last good router on the path or the last router that has a valid return path to the server issuing the traceroute.05Harrison_ch04. The router immediately after the last visible one is the one at which the routing changes. (!H or !N messages may be produced. A more convenient way of doing this is to use a site like traceroute. You may have a routing loop in which packets bounce between two routers and never get to the intended destination. (!H or !N messages may be produced. ☞ Log on to this next hop router. It’s usually good to do the following traceroute ☞ ☞ ☞ ☞ ☞ ☞ Log on to the last visible router.

VIEWING PACKET FLOWS WITH TCPDUMP The tcpdump command is one of the most popular packages for viewing the flow of packets through your Linux box’s NIC card. 192. MTR is automatically installed as part of Fedora Linux. 192.35 0% 16 16 24 23 29 42 5. especially if you are doing simpler types of troubleshooting.Display mode R .” There is even a free Windows version called WinMTR.1 0% 17 17 32 10 15 32 2. making it much more convenient than a regular traceroute.18. worst. You type mtr followed by the target IP address to get output similar to the following: [root@bigboy tmp]# mtr 192. you can download the RPM software installation package from many of the Fedora download sites.qxd 2/25/05 2:35 PM Page 79 Viewing Packet Flows with tcpdump 79 USING MTR TO DETECT NETWORK CONGESTION Matt’s Traceroute is an application you can use to do a repeated traceroute in real time. 192.Quit Packets Pings Hostname %Loss Rcv Snt Last Best Avg Worst 1. The constant updates enable you not only to visually determine which hops are slow.168.168.05Harrison_ch04. but also to determine when they appear to be slow. It is installed by default on Red Hat/Fedora Linux and has very simple syntax.26 0% 16 16 23 21 26 37 ^C [root@bigboy tmp]# One of the nice features of MTR is that it gives you the best.168. 192.52] Bigboy Fri Feb 20 17:19:17 2004 Keys: D . “Installing RPM Software. The advantage of this is that you can let MTR run for an extended period of time.26 Matt’s traceroute [v0.168.1.Restart statistics Q .168. If MTR isn’t installed on your system. The installation of RPMs is covered in Chapter 6.2.3.15 0% 17 17 23 14 18 25 4. 192. acting as a constant monitor of communication path quality.168. .254 0% 17 17 12 11 18 41 3. The constant refreshing of the screen also enables you to instantaneously spot transient changes in quality fairly easily. It is a good tool to use whenever you suspect there is some intermittent network congestion. it dynamically shows the round-trip time to reach each hop along the traceroute path.25.25. and average round-trip times in milliseconds for the probe packets between each hop along the way to the final destination.

View icmp packets. Listen on interface. Don’t print a timestamp at the beginning of each line. tcpdump uses command-line switches to modify the output. . Lack of communication could be due to the following: ☞ Bad routing ☞ Faulty cables. View TCP packets with either a source or destination TCP port of port-number. If this is not specified. You can also add expressions after all the command-line switches. These act as filters to limit the volume of data presented on the screen. Some useful expressions are listed in Table 4.2.qxd 2/25/05 9:56 AM Page 80 80 Simple Network Troubleshooting Chapter 4 One of the most common uses of tcpdump is to determine whether you are getting basic two-way communication. Table 4.05Harrison_ch04.2 Tcpdump Possible tcpdump Switches Command -c -t -i -w -C Switch Description Stop after viewing count packets. View UDP packets with either a source or destination UDP port of port-number. interfaces of devices in the packet flow ☞ The server not listening on the port because the software isn’t installed or started ☞ A network device in the packet path blocking traffic. common culprits are firewalls. Dump the output to a specially formatted tcpdump dump file. You can also use keywords such as and or or between expressions to further fine-tune your selection criteria.3 Tcpdump Useful tcpdump Expressions Description Command Expression host host-address icmp tcp port port-number udp port port-number View packets from the IP address host-address. Some of the more useful command-line switches are listed in Table 4.3. routers with access control lists and even your Linux box running iptables Analyzing tcpdump in much greater detail is beyond the scope of this section. Table 4. tcpdump uses the lowest numbered interface that is UP. Specify the size the dump file must reach before a new one with a numeric extension is created. Like most Linux commands.

sackOK.timestamp 75227931 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# 2013297020:2013297020(0) win 5840 <mss 0.928257 smallfry > bigboy.ssh: S 1460.sackOK. ☞ The second column of data shows the packet source and then the destination IP address or server name of the packet.32938: R 0x10] smallfry.ssh: S 1460.968556 bigboy > smallfry: icmp: echo reply 9 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# In this example: ☞ The first column of data is a packet timestamp.timestamp 75227931 bigboy.928365 bigboy.ssh > smallfry.32938 > bigboy.102 and tcp port 22 tcpdump: listening on wlan0 smallfry.wscale 0> (DF) [tos 0x10] 0:0(0) ack 1 win 0 (DF) [tos 0x10] 2013297020:2013297020(0) win 5840 <mss 0.32938 > bigboy.ssh: S 1460.943926 smallfry > bigboy.102 on TCP port 22 with no timestamps in the output (-t switch): [root@bigboy tmp]# tcpdump -i wlan0 -t host 192.ssh > smallfry.nop.qxd 2/25/05 9:56 AM Page 81 Viewing Packet Flows with tcpdump 81 The following is an example of tcpdump being used to view ICMP packets going through interface wlan0: [root@bigboy tmp]# tcpdump -i wlan0 icmp ping tcpdump: listening on wlan0 21:48:58.05Harrison_ch04.wscale 0> (DF) [tos 0x10] 0:0(0) ack 2013297021 win 0 (DF) [tos 2013297020:2013297020(0) win 5840 <mss 0.927510 bigboy > smallfry: icmp: echo reply 21:48:58.org: icmp: echo request (DF) 21:48:58. > smallfry: icmp: echo reply 21:48:58.32938 > bigboy.nop.1.32938: R smallfry.1. ☞ The third column shows the packet type.927091 smallfry > bigboy: icmp: echo request (DF) 21:48:58.sackOK.963966 bigboy > smallfry: icmp: echo reply 21:48:58.timestamp 75227931 bigboy.my-web-site.168.962244 bigboy > smallfry: icmp: echo reply 21:48:58.my-web-site.org: icmp: echo request (DF) 21:48:58. The following example shows tcpdump being used to view packets on interface wlan0 to/from host 192.wscale 0> (DF) [tos 0x10] . ☞ Two-way communication is occurring as each echo gets an echo reply.944034 bigboy > smallfry: icmp: echo reply 21:48:58.168.nop.

1753 > bigboy-100.ssh > 192-168-1-242.ssh: .019042 IP 192-168-1-242.05Harrison_ch04.myweb-site. not just brief IP information that normally goes to the screen. Without the -n switch: [root@bigboy tmp]# tcpdump -i eth1 tcp port 22 tcpdump: verbose output suppressed. The -n switch stops DNS name lookups and makes tcpdump work more reliably. This can slow down tcpdump so much that it appears not to be working at all.org.1753 > bigboy-100.qxd 2/25/05 9:56 AM Page 82 82 Simple Network Troubleshooting Chapter 4 In this example: ☞ The first column of data shows the packet source and then the destination IP address or server name of the packet. and the capability of automatically highlighting certain error conditions such as data retransmissions: tcpdump -i eth1 -w /tmp/packets. Common Problems with tcpdump By default tcpdump will attempt to determine the DNS names of all the IP addresses it sees while logging data.org.myweb-site. colorization of packet records based on criteria deemed interesting.019118 IP bigboy-100.org.org.1753: P 165:401(236) ack 0 win 6432 . use -v or -vv for full protocol decode listening on eth1.1753: P 1:165(164) ack 0 win 6432 02:24:35. link-type EN10MB (Ethernet).my-web-site.my-web-site.org.myweb-site. This can then be analyzed by graphical analysis tools such as Ethereal.818478 IP bigboy-100. The application is part of the Fedora RPM suite.ssh: .my-web-site.818398 IP 192-168-1-242. but that shouldn’t discourage you from using it. capture size 96 bytes 02:24:34. ack 165 win 65307 02:24:35.org.myweb-site.dump tcp port 22 Covering Ethereal is beyond the scope of this book. ☞ The client named bigboy is using port 32938 to communicate with the server named smallfry on the TCP SSH port 22. The following are examples of how the -n switch affects the output. ☞ Two-way communication is occurring. Analyzing tcpdump files By using the -w filename option you can send the entire Ethernet frame.org. ack 318574223 win 65471 02:24:34. to a file. ☞ The second column shows the TCP flags within the packet.ssh > 192-168-1-242.my-web-site. and a Windows version is also available. with customized filters. which is available in both Windows and Linux.org.

1753: P 165:353(188) ack 0 win 6432 02:25:53.ssh > 192.ssh: P 0:20(20) ack 353 win 64811 02:25:53.ssh > 192. The command-line options and screen output mimic that of tcpdump in many ways.242.org.qxd 2/25/05 9:56 AM Page 83 Viewing Packet Flows with tethereal 83 02:24:35.1753 > 192.168.168.ssh: .176299 IP 192-168-1-242.168.1753: P 401:629(228) ack 20 win 6432 6 packets captured 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# With the -n switch: [root@bigboy tmp]# tcpdump -i eth1 -n tcp port 22 tcpdump: verbose output suppressed.ssh: .242.100. .ssh > 192-168-1-242.242.1.1753: P 353:541(188) ack 20 win 6432 6 packets captured 7 packets received by filter 0 packets dropped by kernel [root@bigboy tmp]# VIEWING PACKET FLOWS WITH TETHEREAL The tethereal program is a text version of the graphical Ethereal product that is part of the Fedora Linux RPM suite.242.068511 IP 192. It can additionally limit the total number of files created before overwriting the first one in the queue.1. use -v or -vv for full protocol decode listening on eth1.1. The tethereal command has the capability of dumping data to a file like tcpdump and creating new files with new filename extensions when a size limit has been reached.4 and 4.168.1.168.1. though the dump file format is identical to tcpdump. capture size 96 bytes 02:25:53.242. but tethereal has a number of advantages.168.269152 IP 192.ssh > 192.168.5 show some popular command switches and expressions that can be used with tethereal.408556 IP 192.242.org.my-web-site.1753 > 192.100.168.100.168.myweb-site.168.1.my-web-site.269205 IP 192.168. ack 165 win 64999 02:25:53.176337 IP bigboy-100.myweb-site.1.068606 IP 192. which is also known as a ring buffer.100.1.1.ssh: P 0:20(20) ack 401 win 65071 02:24:35.1753 > 192.100.1. The tethereal screen output is also more intuitive to read.168.1.1.408589 IP 192.1753 > bigboy-100. ack 318576011 win 65163 02:25:53.100. link-type EN10MB (Ethernet). Tables 4.05Harrison_ch04.org.org.1753: P 1:165(164) ack 0 win 6432 02:25:53.

168. Listen on interface. View UDP packets with either a source or destination UDP port of port-number.100 -> 192.168.603356 192.000000 192.102 -> 192.1.168.168.1. followed by the sequence numbering.168.100. ACK] > 1442 [ACK] > 1442 [FIN. Also notice how the command line mimics that of tcpdump: [root@smallfry tmp]# tethereal -i eth0 tcp port 80 192. but the formatting lines up in neat columns on your screen.1. ACK] > http [ACK] . View TCP packets with packets either a source or destination TCP port of port-number.168. Determine the size of the ring buffer when the -C switch is selected.168.1.1. View icmp packets. tethereal Command Switch -c -i -w -C -b Table 4.168.603211 192.102 -> 192.05Harrison_ch04.100 Capturing on eth0 0.102 TCP http Seq=3291904937 Ack=3325831830 Win=46 Len=0 2.102 -> 192.1. Dump the output to a specially formatted tethereal dump file.102 -> 192. Specify the size the dump file must reach before a new one with a numeric extension is created.1.168.1.168. The tethereal output groups the IP addresses and TCP ports together and then provides the TCP flags.1.168.5 tethereal Command Expression host host-address Icmp tcp port port-number udp port port-number In the next example we’re trying to observe an HTTP (TCP port 80) packet flow between server smallfry at address 192.1.1.168.1.1.1.168.100 TCP 1442 Seq=3325831829 Ack=3291904937 Win=5840 Len=0 2.100 -> 192.1.qxd 2/25/05 9:56 AM Page 84 84 Simple Network Troubleshooting Chapter 4 Table 4.000223 192.100 TCP 1442 Seq=3325831830 Ack=3291904938 Win=5840 Len=0 [root@smallfry tmp]# and host > http [SYN] > 1442 [SYN. tethereal will use the lowest numbered interface that is UP.1.1.603398 192. If this is not specified. Useful tethereal Expressions Description View packets from the IP address host-address.102 TCP http Seq=3291904936 Ack=3325831829 Win=5792 Len=0 0.100 TCP 1442 Seq=3325831829 Ack=3291904937 Win=5840 Len=0 2.000157 192. ACK] > http [ACK] > http [FIN.168.102 TCP http Seq=3291904937 Ack=3325831830 Win=46 Len=0 2.100 -> 192.168.168.102 and bigboy at IP address 192. making analysis much easier.168.602804 192. It may not be apparent on this page.4 Possible tethereal Switches Description Stop after viewing count packets.100 TCP 1442 Seq=3325831828 Ack=0 Win=5840 Len=0 0.

92 returns the Web site named www.92 [root@bigboy tmp]# Using nslookup to Check Your IP Address The nslookup command can operate in the opposite way in which a query against the address 216.com Address: 216.conf file. DNS is the system that helps map an IP address to your Web site’s domain name and your site may suddenly become unavailable if the mapping is incorrect.193... As mentioned before.. as explained in Chapter 18. The nslookup command is very easy to use.linuxhomenetworking. or an incorrect configuration on the DNS server. which is then displayed on the screen.. but that shouldn’t discourage you from using it.com: [root@bigboy tmp]# nslookup 216.linuxhomenetworking.151.193.151.151..in-addr.linuxhomenetworking.com .193.net. .92 .arpa name = extra193-92. BASIC DNS TROUBLESHOOTING Sometimes the source of problems can be due to misconfigured DNS rather than poor network connectivity. you just need to type the command.151.193. Non-authoritative answer: 92.qxd 2/25/05 9:56 AM Page 85 Basic DNS Troubleshooting 85 Using graphical Ethereal to analyze tethereal dump files is beyond the scope of this book.151.. Using nslookup to Test DNS The nslookup command can be used to get the associated IP address for your domain and vice versa. .linuxhomenetworking. Failures can be caused by your server not having the correct value set in the /etc/resolv. Name: www..193. The application is part of the Fedora RPM suite and a Windows version is also available. . followed by the IP address or Web site name you want to query.my-isp-provider.com: the IP address 216..05Harrison_ch04. poor connectivity to your DNS server. The command actually queries your DNS server for a response.92 for the site [root@bigboy tmp]# nslookup www. Using nslookup to Check Your Web Site Name Here we see nslookup returning www.216.

96.151.3 Using domain server: Name: 68.216.com www.my-isp-provider.87..com has address 216. this can be achieved by adding the DNS server’s IP address immediately after the IP address of the Web site name you intend to query: [root@bigboy tmp]# nslookup www.71..216.05Harrison_ch04. . Using the host Command to Test DNS More recent versions of Linux have started to use the host command for basic DNS testing.linuxhomenetworking.151. The Linux host command can be used as a good replacement.151.my-ispprovider.96.96.arpa nameserver = dns2.linuxhomenetworking.arpa nameserver = dns1.linuxhomenetworking.in-addr. [root@bigboy tmp]# [root@bigboy tmp]# host www.92 [root@bigboy tmp]# [root@zippy root]# host www.net.3 Address: 68.1 [root@bigboy tmp]# Using nslookup to Query a Specific DNS Server Sometimes you might want to test the DNS mapping against a specific DNS server.net.my-isp-provider.87.96. dns1.87. syntax is identical to that of nslookup and the resulting output is very similar: [root@bigboy tmp]# host 216.96.87.193.151.arpa domain name pointer extra193-92..com Address: 216. but can still be used with Windows. Server: 68.in-addr.193.34 [root@zippy root]# .3 . Fortunately.com has address 65.linuxhomenetworking.com 68.96.net internet address = 216..193.151.87.net.qxd 2/25/05 9:56 AM Page 86 86 Simple Network Troubleshooting Chapter 4 Authoritative answers can be found from: 193.linuxhomenetworking.in-addr.3#53 Aliases: www.3 Address: 68.115.3#53 Name: www.151.my-isp-provider.193.linuxhomenetworking.216.151.192. 193.92 92.com 68.92 [root@bigboy tmp]# Note The nslookup command will probably be removed from future releases of Linux.87.

1. Whenever you are in doubt.168. It is a favorite tool of malicious surfers and therefore should be used to test external as well as internal servers under your control...168. . but it can be used in a corporate environment to detect vulnerabilities in your network.” Tries to detect the operating system of the remote server based on known responses to various types of packets. you can get a list of available nmap options by just entering the command without arguments at the command prompt: [root@bigboy tmp]# nmap Nmap V.3 seconds with a value of 0 in “paranoid mode. -O -p -s Here is an example of trying to do a scan using valid TCP connections (-sT) in the extremely slow insane mode (-T 5) from ports 1 to 5000: [root@bigboy tmp]# nmap -sT -T 5 -p 1-5000 192. 3. It isn’t usually an important tool in the home environment.6 Argument -P0 -T Commonly Used nmap Options Description Attempts to ping a host before scanning it. This option can be used to send them from 60 seconds apart with a value of 5. Defines the timing between the packets set during a port scan..1..org/nmap/ ) Interesting ports on whoknows.my-site-int.153): (The 4981 ports scanned but not shown below are in state: closed) Port State Service . Some firewalls can detect the arrival of too many nonstandard packets within a predetermined time frame.153 Starting nmap V. such as servers running unauthorized network applications.05Harrison_ch04.6. “insane mode. If the server is being protected from ping queries. [root@bigboy tmp]# Some of the more common nmap options are listed in Table 4.com (192.00 ( www.insecure.qxd 2/25/05 9:56 AM Page 87 Using nmap 87 USING NMAP You can use nmap to determine all the TCP/IP ports on which a remote server is listening.” to 0. Lists the TCP/IP port range to scan. Defines a variety of scan methods that use either packets that comply with the TCP/IP standard or are in violation of it. 3. you can use this option to force it to scan anyway. but you should also refer to the nmap man pages for full descriptions of them all. Table 4.00 Usage: nmap [Scan Type(s)] [Options] <host or net list> Some Common Scan Types (‘*’ options require root privileges) * -sS TCP SYN stealth port scan (default if privileged (root)) -sT TCP connect() port scan (default for unprivileged users) * -sU UDP port scan -sP ping scan (Find any reachable machines) .

Sometimes DNS lookups aren’t enough to determine who owns an offending IP address. such as e-mail SPAM.. Sometimes the attack isn’t in the form of a constant bombardment that your server can’t handle. You need another tool. Here is an example for the yahoo.com +1. a traceroute won’t provide any contact information for the ISP. 701 First Avenue Sunnyvale CA 94089 US domainadmin@yahoo-inc. but you should go the extra mile and purchase a text specifically on Linux security to help protect you against attempts at malicious security breaches.. but of the type that you can’t handle. DETERMINING THE SOURCE OF AN ATTACK Sometimes you realize that your system is under a denial-of-service attack. One of the better ones to use is the whois command.com domain: [root@bigboy tmp]# whois yahoo. ISPs are usually very sensitive to complaints about SPAM.4083493300 .qxd 2/25/05 9:56 AM Page 88 88 Simple Network Troubleshooting Chapter 4 21/tcp 25/tcp 139/tcp 199/tcp 2105/tcp 2301/tcp 3300/tcp open open open open open open open ftp smtp netbios-ssn smux eklogin compaqdiag unknown Nmap run completed -.1 IP address (1 host up) scanned in 8 seconds [root@bigboy tmp]# Full coverage of the possibilities on nmap as a security scanning tool are beyond the scope of this book. Use it with an IP address or DNS domain as its sole argument and it will provide you with all the administrative information you need to start your hunt.. Administrative Contact: Domain Administrator (NIC-1382062) Yahoo! Inc.com . This could be either malicious or simply someone rapidly downloading all the pages of your Web site with the Linux wget command. but though you may have the IP address. Symptoms include a large number of established connections when viewed with the netstat command or an excessive number of entries in your firewall or Web server logs..05Harrison_ch04. .

my-web-site.my-web-site. This makes it easier to track down individuals rather than groups of users. .org: [root@bigboy tmp]# who root pts/0 Jun 19 09:26 (reggae.my-web-site.00:35 (17:10) wtmp begins Sun Jun 15 16:29:18 2003 [root@bigboy tmp]# In this example someone from reggae. which lists the last users who logged into the system. but mostly to figure out who made the mistake that crashed the system or blew up Apache with a typographical error in the httpd. They can then get root privileges by using sudo.05Harrison_ch04. This isn’t just to help track the activities of malicious users.org Wed Jun 18 01:07 .conf file.my-web-site..4.09:26 (1+08:18) reboot system boot 2.. Here are some examples: [root@bigboy tmp]# last -100 root pts/0 reggae.org Thu Jun 19 09:26 still logged in root pts/0 reggae.org Tue Jun 17 21:57 .my-web-site.org logged into bigboy as user root.my-web-site.qxd 2/25/05 9:56 AM Page 89 Who Has Used My System? 89 Fax.. I generally prefer not to give out the root password and let all the systems administrators log in with their own individual logins. The last Command The most common command to determine who has logged into your system is last. Here we see a user logged as root from server reggae.my-web-site.4083493301 . [root@bigboy tmp]# WHO HAS USED MY SYSTEM? It is always important to know who has logged into your Linux box.org) [root@bigboy tmp]# .18-14 Wed Jun 18 01:07 (1+08:21) root pts/0 reggae.+1.. The who Command The who command is used to see who is currently logged in to your computer.org Mon Jun 16 07:24 .down (03:07) root pts/0 reggae.

05Harrison_ch04. especially as network instability is often under the control of network engineers who need evidence pointing to problems in their domain of expertise before they will be convinced to act. The next chapter builds on this new knowledge and expands your troubleshooting skills to include the reading of Linux error log files to assist in the diagnosis of unexpected Linux application behavior. .qxd 2/25/05 9:56 AM Page 90 90 Simple Network Troubleshooting Chapter 4 CONCLUSION One of the greatest sources of frustration for any systems administrator is to try to isolate whether poor server performance is due to a network issue or problems with an application or database. The worry can be amplified. These tips should help provide you with a definitive answer by enabling you to isolate the source of most network problems and helping you make their resolution much faster.

each with their own configuration files and help pages. but it also makes Linux system administration daunting. syslog and the logrotate utility that cleans up log files are both relatively easy to configure but they frequently don’t get their fair share of coverage in most texts. 91 . but will usually log them somewhere. This can be invaluable in correlating the timing and causes of related events on your system. It is also important to know that applications frequently don’t display errors on the screen. Knowing the precise message that accompanies an error can be vital in researching malfunctions in product manuals. and Web searches.qxd 2/25/05 9:57 AM Page 91 C H A P T E R 5 Troubleshooting Linux with syslog IN THIS CHAPTER ☞ syslog ☞ logrotate ☞ Conclusion There are hundreds of Linux applications on the market.06Harrison_ch05. in most cases. Linux applications use the syslog utility to export all their errors and status messages to files located in the /var/log directory. online documentation. I’ve included syslog here as a dedicated chapter to both emphasize its importance to your Linux knowledge and prepare you with a valuable skill that will help you troubleshoot all the Linux various applications that will be presented throughout the book. This variety makes Linux vibrant. Fortunately.

conf configuration file to place messages of differing severity and facilities in different files.none. and the second lists the files to which they should be logged. This procedure will be covered next.none /var/log/messages In this case. By default. ☞ The second describes the degree of severity of the message.conf configuration file. Here is a sample: . Red Hat/Fedora’s /etc/ syslog. You can make this logging even more sensitive by replacing the line above with one .06Harrison_ch05.qxd 2/25/05 2:36 PM Page 92 92 Troubleshooting Linux with syslog Chapter 5 SYSLOG syslog is a utility for tracking and logging all manner of system messages from the merely informational to the extremely critical.authpriv. Table 5.cron.info.conf File The files to which syslog writes each type of message received is set in the /etc/syslog. This file consists of two columns: The first lists the facilities and severity of messages to expect.1 syslog Facilities Severity Level 0 1 2 3 4 5 6 7 Keyword emergencies alerts critical errors warnings notifications informational debugging Description System unusable Immediate action required Critical condition Error conditions Warning conditions Normal but significant conditions Informational messages Debugging messages The /etc/syslog. or authentication facilities/subsystems.1.none. There are eight in all and they are listed in Table 5. Each system message sent to the syslog server has two descriptive labels associated with it that makes the message easier to handle: ☞ The first describes the function (facility) of the application that generated it.mail. applications such as mail and cron generate messages with easily identifiable facilities named mail and cron. all messages of severity “info” and above are logged. You can configure syslog’s /etc/syslog. For example. cron. but none from the mail.conf file is configured to put most of the messages in the file /var/log/messages.

debug /var/log/ /var/log/messages Certain applications will additionally log to their own application specific log files and directories independent of the syslog. This may be more suitable for troubleshooting: *.conf file. Issue this command to do so: [root@bigboy tmp]# service syslog restart How to View New Log Entries as They Happen If you want to get new log entries to scroll on the screen as they occur.conf file to be safe.qxd 2/25/05 9:57 AM Page 93 syslog 93 that captures all messages from debug severity and above in the messages file. you can use this command: [root@bigboy tmp]# tail -f /var/log/messages . Here are some common examples: Files /var/log/maillog /var/log/httpd/access_log : Mail : Apache web server page access logs Directories /var/log /var/log/samba /var/log/mrtg /var/log/httpd : Samba messages : MRTG messages : Apache webserver messages Note In some older versions of Linux the /etc/syslog.06Harrison_ch05. Check the formatting of your /etc/syslog.conf file was very sensitive to spaces and would recognize only tabs. The use of spaces in the file would cause unpredictable results. Activating Changes to the syslog Configuration File Changes to /etc/syslog.conf will not take effect until you restart syslog.

it becomes easier to correlate events across your company. With all servers logging to a central syslog server. and # once for processing with ‘ksymoops’ # -x disables all klogd processing of oops messages entirely # See klogd(8) for more details KLOGD_OPTIONS=”-2” . As we saw previously. Here is an example: [root@bigboy tmp]# more /var/log/messages Logging syslog Messages to a Remote Linux Server Logging your system messages to a remote server is a good security practice. It also checks the file /etc/sysconfig/syslog to determine the various modes in which it should operate. you can pipe it through the more command so that you only get one screen at a time. especially if you restrict the user access to the logging server. grep will help you search for all occurrences of a string in a log file. syslog will not listen for remote messages unless the SYSLOGD_OPTIONS variable in this file has an -r included in it: # Options to syslogd # -m 0 disables ‘MARK’ messages. once for klogd to decode. syslog checks its /etc/syslog. syslog doesn’t expect to receive messages from remote clients. Here is an example: [root@bigboy tmp]# grep string /var/log/messages | more You can also just use the plain old more command to see one screen at a time of the entire log file without filtering with grep.qxd 2/25/05 9:57 AM Page 94 94 Troubleshooting Linux with syslog Chapter 5 Similar commands can be applied to all log files. Configuring the Linux syslog Server By default. This is probably one of the best troubleshooting tools available in Linux. Another good command to use apart from tail is grep. # -r enables logging from remote machines # -x disables DNS lookups on messages received with -r # See syslogd(8) for more details SYSLOGD_OPTIONS=”-m 0 -r” # Options to klogd # -2 prints all kernel oops messages twice.conf file to determine the expected names and locations of the log files it should create.06Harrison_ch05. It also makes covering up mistakes or malicious activities harder because the purposeful deletion of log files on a server cannot simultaneously occur on your logging server. Here’s how to configure your Linux server to start listening for these messages.

debug @loghost /var/log/messages You have now configured all debug messages and higher to be logged to both server bigboy (“loghost”) and the local file /var/log/messages. Add an entry in the /etc/hosts file in the format: IP-address fully-qualified-domain-name hostname “loghost” Example: 192.0:* [root@bigboy tmp]# Configuring the Linux Client The syslog server is now expecting to receive syslog messages.168. which you can verify using either one of the following netstat command variations: [root@bigboy tmp]# netstat -a | grep syslog udp 0 0 *:syslog *:* [root@bigboy tmp]# netstat -an | grep 514 udp 0 0 0. This is done by editing the /etc/hosts file on the Linux client named smallfry. Here are the steps: 1.org bigboy loghost Now your /etc/hosts file has a nickname of “loghost” for server bigboy.conf file to make the syslog messages get sent to your new loghost nickname: *.my-web-site.100 bigboy.0:514 0.0. The next thing you need to do is edit your /etc/syslog.qxd 2/25/05 9:57 AM Page 95 syslog 95 You have to restart syslog on the server for the changes to take effect. Linux Client [root@smallfry tmp]# service lpd restart Stopping lpd: [ OK ] Starting lpd: [ OK ] [root@smallfry tmp]# .1.0. 2. The server will now start to listen on UDP port 514. 3. such as restarting the lpd printer daemon and making sure the remote server sees the messages.06Harrison_ch05.0. Remember to restart syslog to get the remote logging started.debug *. Determine the IP address and fully qualified hostname of your remote logging host.0. You can now test to make sure that the syslog server is receiving the messages with a simple test. You have to configure your remote Linux client to send messages to it.

and LocalDirectors. Appendix IV has examples of how to configure syslog to do this with Cisco devices using separate log files for the routers. The /etc/logrotate. Apr 11 22:09:35 smallfry lpd: lpd shutdown succeeded Apr 11 22:09:39 smallfry lpd: lpd startup succeeded . syslog syslog and Firewalls syslog listens by default on UDP port 514. In the case below.. [root@bigboy tmp]# syslog Configuration and Cisco Network Devices reserves facilities local0 through local7 for log messages received from remote servers and network devices. ☞ The rotate parameter specifies the number of copies of log files logrotate will maintain.06Harrison_ch05. ☞ The create parameter creates a new log file after each rotation. while allowing 7 copies. LOGROTATE The Linux utility logrotate renames and reuses system error log files on a periodic basis so that they don’t occupy excessive disk space. and load balancers—each logging with a different facility—can each have their own log files for easy troubleshooting. switches.conf file is logrotate’s general configuration file in which you can specify the frequency with which the files are reused: ☞ You can specify either a weekly or daily rotation parameter. In the case below. .. PIX firewalls. switches. the 4 copy option is commented out with a #..conf File The /etc/logrotate. If you are logging to a remote syslog server via a firewall. . you have to allow traffic on this port to pass through the security device.. CSS load balancers.. syslog messages usually have UDP port 514 for both their source and destination UDP ports.qxd 2/25/05 9:57 AM Page 96 96 Troubleshooting Linux with syslog Chapter 5 Linux Server [root@bigboy tmp]# tail /var/log/messages . . Routers. the weekly option is commented out with a #. allowing daily updates.... firewalls.

qxd 2/25/05 9:57 AM Page 97 logrotate 97 Therefore.06Harrison_ch05.3 logfile. Here are some sample files that define the specific files to be rotated for each application.pid 2> /dev/null` 2> /dev/null || true endscript } . The files will have the following names.d/syslog File (for General System Logging) /var/log/messages /var/log/secure /var/log/maillog /var/log/spooler /var/log/boot.1 logfile. with logfile the current active version: logfile logfile.2 logfile. The /etc/logrotate.4 logfile.d Directory Most Linux applications that use syslog put an additional configuration file in this directory to specify the names of the log files to be rotated.log /var/log/cron { sharedscripts postrotate /bin/kill -HUP `cat /var/run/syslogd. our sample configuration file will create daily archives of all the logfiles and store them for seven days.5 logfile.conf # rotate log files weekly #weekly # rotate log files daily daily # keep 4 weeks worth of backlogs #rotate 4 # keep 7 days worth of backlogs rotate 7 # create new (empty) log files after rotating old ones create The /etc/logrotate.6 Sample Contents of /etc/logrotate. It is a good practice to verify that all new applications that you want to use the syslog log have configuration files in this directory.0 logfile.

issue the logrotate command with just that filename as the argument: [root@bigboy tmp]# logrotate -f /etc/logrotate.06Harrison_ch05.qxd 2/25/05 9:57 AM Page 98 98 Troubleshooting Linux with syslog Chapter 5 The /etc/logrotate. and not all of them.pid 2> /dev/null` 2> /dev/null || true endscript } Activating logrotate The logrotate settings in the last section will not take effect until you issue the following command: [root@bigboy tmp]# logrotate -f If you want logrotate to reload only a specific configuration file.d/syslog Compressing Your Log Files On busy Web sites the size of your log files can become quite large. Compression can be activated by editing the logrotate.conf file and adding the compress option.log { notifempty missingok sharedscripts copytruncate postrotate /bin/kill -HUP `cat /var/lock/samba/*.d/apache File (for Apache) /var/log/httpd/access_log /var/log/httpd/agent_log /var/log/httpd/error_log /var/log/httpd/referer_log { missingok sharedscripts postrotate /bin/kill -HUP `cat /var/run/httpd.pid 2>/dev/null` 2> /dev/null || true endscript } The /etc/logrotate.d/samba File (for Samba) /var/log/samba/*. .

gz /var/log/messages.info. the Apache Web server.gz /var/log/messages..6.34]. Linux mail.gz /var/log/messages.conf # # Activate log compression compress The log files will then start to become archived with the gzip utility. relay=www.info. which can be configured to ignore and then log all undesirable packets.gz /var/log/messages. [222.gz /var/log/messages. mailer=esmtp.1.0.info>.clematis4spiders.5.gz .. pri=6388168. and finally. delay=2+20:45:44. dsn=4.7. xdelay=00:00:02. each file having a ..1.06Harrison_ch05. This syslog chapter should make you more confident to learn more about these applications via experimentation because you’ll at least know where to look at the first sign of trouble.66. [root@bigboy tmp]# CONCLUSION In the next chapter we cover the installation of Linux applications.3..gz /var/log/messages. Use the command with the compressed file’s name as the argument: [root@bigboy tmp]# zcat /var/log/messages.0.gz [root@bigboy tmp]# Viewing the contents of the files still remains easy because the zcat command can quickly output the contents to the screen. and the use of syslog will become increasingly important especially in the troubleshooting of Linux-based firewalls. .gz extension. whose configuration files are probably the most frequently edited system documents of all and which correspondingly suffer from the most mistakes. [root@bigboy tmp]# ls /var/log/messages* /var/log/messages /var/log/messages. Nov 15 04:08:02 bigboy httpd: httpd shutdown succeeded Nov 15 04:08:04 bigboy httpd: httpd startup succeeded Nov 15 04:08:05 bigboy sendmail[6003]: iACFMLHZ023165: to=<tvaughan@clematis4spiders. stat=Deferred: Connection refused by www. . which logs all application programming errors generated by some of the popular scripting languages such as PERL and PHP.4.134.2.qxd 2/25/05 9:57 AM Page 99 Conclusion 99 # # File: /etc/logrotate.clematis4spiders.

qxd 2/25/05 9:57 AM Page 100 .06Harrison_ch05.

qxd 2/25/05 9:58 AM Page 101 C H A P T E R 6 Installing RPM Software IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Where to Get Commonly Used RPMs Getting RPMs Using Web-Based FTP Getting RPMs Using Command-Line Anonymous FTP Getting RPMs Using wget Automatic Updates with yum How to Install RPMs Manually How to Install Source RPMs RPM Installation Errors How to List Installed RPMs Listing Files Associated with RPMs Uninstalling RPMs Which RPMs Start at Boot Time? Installing Software Using tar Files Conclusion You’ll frequently need to install additional software on your Linux server that you didn’t think you’d need when you first installed the operating system. Regular RPM package files are used for installations in which the kernel. or master program. This is the usual scenario for most departmental servers. This could be because of new business requirements for additional packages or the need to install new administrative tools to make your job easier.07Harrison_ch06. The 101 . Fedora Linux software is primarily available in Red Hat Package Manager (RPM) files. Source RPMs are used when the kernel has been customized to add or drop support selectively for various devices or features for the sake of performance or functionality. hasn’t been customized in any way.

fedora. If you’re new to FTP. Remember that Fedora RPMs may not work on Red Hat operating systems. Software developers who want to use a universally recognizable file format across all flavors of Linux also will make their products available as TAR packages.red site. We discuss using this method in more detail later. RPMs on Your Installation CDs Installing from your distribution CDs is usually easier than having to download files from a remote Web site. or the wget utility. Using FTP or Your Web Browser You can download RPMs from Fedora using the following link: http://download. Start your search in the /pub/fedora/linux/core/ directory and move down the directory tree. and automated downloads using yum.redhat. This makes life easier for the software developer who wrote the package as he or she now has only to create a single package to support all types of customizations. or by using the automatic yum utility.qxd 2/25/05 9:58 AM Page 102 102 Installing RPM Software Chapter 6 procedure for installing source RPMs involves recompiling source code to fit the needs of these kernel customizations. but also devotes a small section on TAR packages near the end because they are still very important. TAR packages are generally more difficult to work with than RPM packages because the archived files within them may or may not need to be compiled and the commands to install the software may vary from package to package. but they are never up to date for very long. Each of these methods is introduced here.com/pub/fedora/linux/core/ hat.07Harrison_ch06. File Transfer Protocol (FTP) client. RPMs Downloaded from Fedora The two most common ways of getting RPMs from Fedora are by manually using FTP or a Web browser. WHERE TO GET COMMONLY USED RPMS There are three commonly used sources for RPMs: Fedora distribution CDs. Both package types use standardized commands for installing the software contained inside. it’s explained later. but is covered in greater detail in sections to follow.fedora. which is the format of choice for Fedora Linux software. . don’t worry. Instructions are usually contained within a file inside the TAR package to help guide the installation. This chapter focuses on the RPM format.com You also can use FTP to download Fedora from the download. making RPMs relatively easy to use. RPMs manually downloaded via a browser.

two of the most popular are the Fedora Web site and rpmfind.net Red Hat and Fedora have only their approved software on their sites. RPMs Downloaded from rpmfind. It enables you to keep the versions of your software up to date by downloading the required packages with the option of installing them afterward. Click on the dhcp-3. Go to the pub/fedora/linux/core/2/i386/os/Fedora/RPMS directory. 2.07Harrison_ch06.qxd 2/25/05 9:58 AM Page 103 Getting RPMs Using Web-Based FTP 103 Using yum The yum program is installed on Fedora systems by default. which is explained later. This site offers packages for a wide variety of software in various formats for many of the popular versions of Linux.net. but different enough to be treated separately. The methods used to get the software from either site are simple. 3. GETTING RPMS USING WEB-BASED FTP There are numerous Web sites from which you can download Fedora RPM software. Installing an incorrect version could adversely affect your system.i386. Note Always remember to select the RPM that matches your version of Linux.net. Let’s say you are running Fedora Core 2 and need to download an RPM for the DHCP server application.0pl2-6. you can select either of these methods in the following sections to get the software. Use your Web browser to go to the Fedora link given earlier. Save the file to your hard drive. Most downloads can be done using HTTP by either clicking on the link or using the wget command. This is discussed in more detail in a later section.rpmfind. Using the rpmfind Web site To use the rpmfind Web site: .rpm link.16. A good general purpose source for additional software is www. 4. It has an easy-to-use search function that enables you to locate the software you need quickly. Using the Fedora Web site To use the Fedora Web site: 1.

ftp> 2. 1. GETTING RPMS USING COMMAND-LINE ANONYMOUS FTP The Web-based method in the last section transparently uses anonymous FTP.com:root): anonymous 331 Please specify the password. or something similar. Name (download. This way anyone can access the data.qxd 2/25/05 9:58 AM Page 104 104 Installing RPM Software Chapter 6 Go to the rpmfind link. drwxr-xr-x 3 ftp ftp 4096 Oct 29 15:59 pub 226 Directory send OK. 1.187.fedora.fedora. First. Password: 230 Login successful.232. Click the link. ftp> pwd 257 “/” ftp> ls 227 Entering Passive Mode (66.232. Have fun. Let’s illustrate this with an example of using anonymous FTP to download the SSH package from download.57.07Harrison_ch06.35). Using binary mode to transfer files.com [root@bigboy tmp]# ftp download. 4. we can use the we have at our disposal: help command to see what options . Type dhcp in the Search box. All transfers are logged..redhat.com Trying 66. 2.187.com (66.fedora.redhat. The right column has links with the actual names of the rpm files. Anonymous FTP enables you to log in and download files from a FTP server using the username anonymous or the shorter username ftp and a password that matches your e-mail address.187. so you can find them later.35.redhat. 6. Connected to download.fedora.155) 150 Here comes the directory listing.232. Save the file to the Linux box’s hard drive. Scroll down for the RPM that matches your version of Fedora.redhat.. It is best to download RPMs to a directory named RPM. 3.redhat. 5.com.35. 220 Fedora FTP server ready. After we’ve logged in. Click the Search button.fedora. we issue the FTP command targeting at the command line: download.

Get a file from the FTP server. Command binary cd dir exit get lcd ls mget mput passive put pwd 3. Bye bye. Same as get. Make the file transfer passive mode. but you can use wildcards like *. Same as put.1. Change the directory on the local machine.07Harrison_ch06.1 FTP Commands Description Copy files in binary mode. List the names of the files in the current remote directory. Put a file from the local machine onto the FTP server. I know that the Fedora Core 2 RPMs are located in the pub/fedora/linux/core/2/i386/os/Fedora/RPMS/ directory and will use the cd command to change my directory to there. Same as dir. We can use the ls command to get a listing of files in this directory: . but you can use wildcards like *. By using the Web browsing feature on the Web site ahead of time. Commands are: ! $ account append ascii bell binary bye case cd cdup chmod close cr Delete ftp> Debug dir disconnect exit form get glob hash help idle image lcd ls macdef Mdelete mdir mget mkdir mls mode modtime mput newer nmap nlist ntrans open prompt passive proxy sendport put pwd quit quote recv reget rstatus rhelp rename reset restart rmdir runique Send Site size status struct system sunique tenex tick trace type user umask verbose ? The commands you’ll most likely use are listed in Table 6. Give the directory name on the local machine. Change directory on the FTP server.qxd 2/25/05 9:58 AM Page 105 Getting RPMs Using Command-Line Anonymous FTP 105 ftp> help Commands may be abbreviated. Table 6.

58.i386..1p2-34.232..07Harrison_ch06.i386.i386. 184281 Oct 28 23:29 openssh-3.6.rpm remote: openssh-3. Next we get the file we need and place it in the local directory /usr/rpm.35. 184281 bytes received in 3. ftp> lcd /usr/rpm Local directory now /usr/rpm ftp> get openssh-3. Finally we use the exit command to leave FTP: ftp> exit 221 Goodbye..6.. -rw-r--r-. ..232.rpm .187.187.qxd 2/25/05 9:58 AM Page 106 106 Installing RPM Software Chapter 6 ftp> cd pub/fedora/linux/core/2/i386/os/Fedora/RPMS/ 250 Directory successfully changed. ############################################################## ############################################################## ####################################################### 226 File send OK.6* mget openssh-3.. root@bigboy tmp]# .3) 150 Here comes the directory listing. In the next example.1p2-34.i386...6.rpm local: openssh-3. ftp> ls open* 227 Entering Passive Mode (66.35. we just aborted this download by typing n.rpm? n ftp> 5..rpm (184281 bytes).1p234.. . ftp> mget openssh-3.1p234.58.rpm 227 Entering Passive Mode (66.25) 150 Opening BINARY mode data connection for openssh-3.i386. The hash command prints # hash signs on the screen during the download: ftp> hash Hash mark printing on (1024 bytes/hash mark).41 secs (53 Kbytes/sec) ftp> Note You can also use wildcards to download the RPMs you need using the mget command.. . .6.6. 226 Directory send OK..1p234. You’ll be prompted for each of the matching RPM files.6. ftp> 4.1p2-34.i386.

. ==> CWD /pub/mirrors/fedora/linux/core/2/i386/os/Fedora/RPMS .5’ saved [529890] [root@bigboy tmp]# AUTOMATIC UPDATES WITH YUM The yum automatic RPM update program comes as a standard feature of Fedora Core. right-click on the desired link.07Harrison_ch06.0pl2-6.edu. Here is an example downloading a DHCP update from Fedora: [root@bigboy tmp]# wget http://linux.stanford.12 KB/s) .18]:21. done.5’ Resolving linux. After you have done this.. ==> TYPE I ..i386.16.rpm.rpm. done.16.0pl2-6.edu/pub/mirrors/fedora/linux/core/2/i386/os/ Fedora/RPMS/dhcp-3..stanford.stanford.i386. ..rpm --17:38:36-ftp://linux.2.0pl2-6.0pl2-6. Connecting to linux. done. ☞ yum automatically figures out not only the RPM packages that need updating. done.. connected. done.edu[171. Logged in! ==> SYST . This provides the added advantage of choosing the most reliable sites in your part of the globe. ==> PWD .0pl2-6..edu/pub/mirrors/fedora/linux/core/2/i386/os/ Fedora/RPMS/dhcp-3. This is especially convenient if you are logged into your Linux box from another machine running a Web browser. Logging in as anonymous .12K/s ETA 17:38:36 (889.i386.qxd 2/25/05 9:58 AM Page 107 Automatic Updates with yum 107 GETTING RPMS USING WGET The wget command can be used to download files quickly when you already know the URL at which the RPM is located.66. You can browse the download site for the RPM you need.i386.. done. Length: 529.16.16. but also all the supporting RPMs..stanford. done.. It then installs them all..i386.. and select copy shortcut (Windows) or Copy Link Location (Linux)...890 (unauthoritative) 100%[===============================>] 529.16. It has a number of valuable features: ☞ You can configure the URLs of download sites you want to use. ☞ yum makes multiple attempts to download RPMs before failing..890 00:00 889.. ==> RETR dhcp-3. ==> PASV ..rpm .rpm => `dhcp-3. you can then select your SSH/telnet/Linux Terminal login window and type the command wget URL.`dhcp-3..

2: Table 6.conf Description Contains logging and fault-tolerance parameters.qxd 2/25/05 9:58 AM Page 108 108 Installing RPM Software Chapter 6 Note Updating packages could cause programs written by you to stop functioning. or it won’t work.html Web site to get a listing of alternative download sites.log pkgpolicy=newest distroverpkg=fedora-release tolerant=1 exactarch=1 . Browse the sites to find the correct locations of the files: ☞ The baseurl URL for [base] would be architecture-type/os subdirectory of your there is a headers subdirectory here.2 Section [main] [base] [updates-released] /etc/yum.conf file. File Format—yum.redhat.com/download/mirrors. Make sure won’t work. ☞ The baseurl URL for [updates-released] would be that of the updates/fedora-version/architecture-type subdirectory of your version of Fedora. There will be RPMs in this subdirectory. especially if they rely on the older version’s features or syntax. There will not be RPMs in this subdirectory.conf file to update Fedora from one of the mirror [main] cachedir=/var/cache/yum debuglevel=2 logfile=/var/log/yum.07Harrison_ch06. or it that of the fedora-version/ version of Fedora. which can usually be left alone Contains the URL (ftp:// or http://) of a mirror site that contains the Fedora base configuration RPMs Contains the URL (ftp:// or http://) of a mirror site that contains updated Fedora RPMs The easiest way to determine the exact URLs to use in the baseurl parameters of the [base] and [updates-released] sections of the file is to go to the http://fedora. Configuring yum All the configuration parameters for yum are stored in the The three basic sections are listed in Table 6. Make sure there is a headers subdirectory here. Here is a sample sites: yum.

FTP firewall rules are more difficult to implement than HTTP. If you use multiple baseurl statements in each section. yum Note It is probably best to select yum update sites that use HTTP instead of FTP. There are a number of reasons for this.qxd 2/25/05 9:58 AM Page 109 Automatic Updates with yum 109 [base] name=Fedora Core $releasever . The $releasever variable refers to the current version of Fedora Core running on your server.com/fedora/core/$releasever/$basearch/ os/ [updates-released] name=Fedora Core $releasever . so your network load should be minimal on average with an update once or twice a week per server. The problem is timing.$basearch .xmission. A small desktop PC with about five to six gigabytes of free disk space per distribution should be sufficient to start with for a dedicated small business yum server. which is determined automatically.07Harrison_ch06. and they are updated infrequently. outbound HTTP access to the Internet is often already allowed in offices. Large RPMs are about twenty-five megabytes in size.$basearch . unlike FTP servers. frequently only selecting the last one in the list: baseurl=url://server1/path/to/files/ url://server2/path/to/files/ url://server3/path/to/files/ Creating Your Own yum Server An obvious advantage of using yum is that you can use it to update a yum server at your office with the same directory structure of the mirror download sites on the Fedora Web site.Base baseurl=http://mirrors.com/fedora/core/updates/$releasever/ $basearch/ Note accepts the use of variables in the configuration file. There is a yum script file in the /etc/cron.xmission. which often have limits on the number of user logins. Note You can list multiple URLs in a baseurl statement like this and yum will try them all. and the $basearch variable maps to the base architecture of your server.daily directory that runs as a cron job . and Web servers are less likely to have connection limits imposed on them.Released Updates baseurl=http://mirrors. yum may act strangely.

If you need help in setting these up. Keeping Your System Current with yum You can make the installed RPM packages on your system up to date with the latest patches using the yum update command.conf file.” discusses Linux FTP servers and Chapter 20.m. you need to enable the viewing of directory structures so that it will be easy for someone to use his or her Web browser to navigate down the directories and double-check the location of the yum files. Your yum server could get overwhelmed with simultaneous update requests from all your yum clients. you can upgrade the server. For most small businesses/departments this should not be a major concern. the yum utility checks the downloaded RPMs against checksum files to help protect against file corruption and malicious forgeries. yum clients can access the yum server using either FTP or HTTP requests. MRTG monitoring is covered in Chapter 22. Finally. Note When setting up an HTTP-based yum server. When the value is set to 1. when set to 0. you could move this script to another location and schedule it as a cron job for different times for each server. Before You Start As of Fedora Core version 3. Chapter 15.” When established. if these measures don’t work. and you can use MRTG on the server to get trend data for its network load. Please refer to the section titled “Signature Keys” later in the chapter before proceeding. “The Apache Web Server. which will significantly reduce your Internet congestion and the associated bandwidth costs. then checks are done.07Harrison_ch06.conf # gpgcheck=1 This is a valuable feature to have but you need to load the checksum files in order for yum to work properly. If the load gets too high. “Monitoring Server Performance.qxd 2/25/05 9:58 AM Page 110 110 Installing RPM Software Chapter 6 every day at 4:00 a. they are disabled: # # File: /etc/yum.” covers the Apache Web server for HTTP requests. you can then configure all your Fedora servers to use this local yum server for all updates. This is set using the gpgcheck variable in the /etc/yum. When used without listing any . You can also consider throttling the NIC card of the yum server to 10Mbps as another interim means of reducing the problem. “Linux FTP Server Setup.

It is always advisable to use yum after installing Linux to make sure the latest versions of software are installed for the sake of improved security and functionality.0. The yum update command updates only a particular RPM package. . Completing update for pango .2-7.127/127 Kernel Updated/Installed.2166.6-4.3.0-55.nptl.FC1.0.FC1. ..2166..22-1.. ..qxd 2/25/05 9:58 AM Page 111 Automatic Updates with yum 111 package-name packages afterward. ..14... Running test transaction: Test transaction complete.0. .rpm samba-client-3. yum will attempt to update them all.i386 .2.nptl.0. Here is an example of output with yum updating your system: [root@bigboy tmp]# yum update Gathering header information file(s) from server(s) Server: Fedora Core 2 .07Harrison_ch06. .90.14.i386 XFree86-Mesa-libGLU 4.Base Server: Fedora Core 2 ..i386] .2-7.i686] [update: samba-client 3.FC1.i386 samba-client 3.22-1....i686 Updated: pango 1.2-7.FC1.Released Updates Finding updated packages Downloading needed headers Resolving dependencies Dependencies resolved I will do the following: [install: kernel 2.66/127 Completing update for binutils ..i386 . Is this ok [y/N]: y Getting samba-client-3..90.6-4..making this kernel the default Installed: kernel 2. . checking for bootloader Grub found ..i386 binutils 2.65/127 Completing update for samba-client .0.. 100% |=========================| 128 kB 05:01 ......i386.. Success! glibc-common 100 % done 1/127 glibc 100 % done 2/127 Stopping sshd:[ OK ] Starting sshd:[ OK ] bash 100 % done 3/127 mozilla-nspr 100 % done 4/127 sed 100 % done 5/127 ...0.i386 initscripts [root@bigboy tmp]# .i386] [update: binutils 2.4.67/127 . . Completing update for XFree86-font-utils .5-4.4.2-7.

Please wait.1.i386 0:5. not adding again Setting up Install Process Setting up Repo: base repomd. The list must be separated by spaces.2-11 set to be installed --> Running transaction check Dependencies Resolved Transaction Listing: Install: net-snmp-utils.1.i386 0:5.1. In this case [root@bigboy tmp]# yum -y install net-snmp-utils Repository updates-released already added. Example of a yum Package Installation Here is a sample installation of an individual package using the RPM installed is the net-snmp-utils package: yum.xml.2-11.2-11 Downloading Packages: net-snmp-utils-5. ---> Package net-snmp-utils. Kernel RPMs may be one of the first sets to go on this list. not adding again Repository base already added. use the yum with the -y switch.1.07Harrison_ch06.2 MB Running Transaction Test Finished Transaction Test Transaction Test Succeeded Running Transaction Installing: net-snmp-utils 100 % done 1/1 Installed: net-snmp-utils.i386 0:5.xml 100% |=========================| 951 B 00:00 Reading repository metadata in from local files base : ############################################ 2622/2622 primary. as in this example: [main] exclude=kernel .xml 100% |=========================| 1.2-11 Complete! [root@bigboy tmp]# 00:48 Remember the following facts about yum: ☞ You can place a list of packages you never want automatically updated in the [main] section.qxd 2/25/05 9:58 AM Page 112 112 Installing RPM Software Chapter 6 Note If you don’t want to be prompted to install the files.1 kB 00:00 Setting up Repo: updates-released repomd.gz 100% |=========================| 88 kB 00:00 MD Read : ################################################## 229/229 updates-re: ################################################## 229/229 Resolving Dependencies --> Populating transaction set with selected packages.i 100% |===================| 6.

58-9. such as a CD-ROM drive. The first method is by using a file previously downloaded to your hard drive.qxd 2/25/05 9:58 AM Page 113 How to Install RPMs Manually 113 does its updates using TCP port 80 for http:// update URLs and uses passive FTP for ftp:// update URLs in /etc/yum. The procedure is as follows: 1.daily/. ☞ More details on configuring yum can be obtained by running the man yum...conf command.rpm Preparing. ☞ Don’t limit yourself to the default yum. and the -v qualifier prints verbose status messages while the command is run.23. such as /tmp. yum ☞ HOW TO INSTALL RPMS MANUALLY There are generally two ways to install RPM files manually.i386. Here is an example of a typical RPM installation command to install the MySQL server package: [root@bigboy tmp]# rpm -Uvh mysql-server-3. Your RPMs will then be located in the CD-ROM’s Fedora/RPMs subdirectory. The cron file is located in /etc/cron. This will have importance for your firewall rules. and then install the RPM: /mnt/cdrom/Fedora/RPMS direc- .conf.07Harrison_ch06. The next step is to issue the rpm -Uvh command to install the package. check the files in the tory. Insert the CD-ROM. and the other is to install the RPM from some sort of removable media.conf URLs because they can become overloaded with requests and make yum perform poorly. The -U qualifier is used for updating an RPM to the latest version. Using Downloaded Files Download the RPMs (which usually have a file extension ending with .rpm) into a temporary directory. the -h qualifier gives a list of hash # characters during the installation. ####################### [100%] 1:mysql-server ####################### [100%] [root@bigboy tmp]# Using CD-ROMs The underlying steps to install RPMs from CDs are similar to those used when installing from your hard disk. ☞ yum runs automatically each day. The main difference is that you have to access your CD-ROM drive by mounting it first to the mnt/cdrom directory.

src.rpm ☞ Here is an example in which we install the TACACS plus package: [root@bigboy rpm]# rpmbuild --rebuild tac_plus-4.0.3-2. .0.3.rpm Preparing.rpm Executing(%prep): /bin/sh -e /var/tmp/rpm-tmp.rpm).. + umask 022 + cd /usr/src/redhat/BUILD + rm -rf tac_plus-4.3-2. This requires you to use source RPM files: ☞ Download the source RPMs or locate them on your CD collection.07Harrison_ch06. They usually have a file extension ending with (.src.0.61594 + umask 022 + cd /usr/src/redhat/BUILD + cd /usr/src/redhat/BUILD + rm -rf tac_plus-4.rpm Installing tac_plus-4..3 + exit 0 [root@bigboy rpm]# .rpm [root@bigboy RPMS]# rpm -Uvh filename..src. ☞ Run the following commands as root. ####################### [100%] 1: filename ####################### [100%] [root@bigboy RPMS]# 2..src. . ☞ Compiling and installing source RPMs with Fedora can be done simply with the rpmbuild command: [root@bigboy tmp]# rpmbuild --rebuild filename. eject the CD-ROM: [root@bigboy RPMS]# cd /tmp [root@bigboy tmp]# eject cdrom [root@bigboy tmp]# Note You can use the rpm command’s --aid switch to make it search the CD-ROM for any other RPM dependencies and install them automatically.0...qxd 2/25/05 9:58 AM Page 114 114 Installing RPM Software Chapter 6 [root@bigboy tmp]# mount /mnt/cdrom [root@bigboy tmp]# cd /mnt/cdrom/Fedora/RPMS [root@bigboy RPMS]# ls filename* filename.3 + /usr/bin/gzip -dc /usr/src/redhat/SOURCES/tac_plus-4.. HOW TO INSTALL SOURCE RPMS Sometimes the packages you want to install need to be compiled in order to match your kernel version..0.tgz + tar -xvvf . When finished.

which really means that a prerequisite RPM needs to be installed.07Harrison_ch06.23. RPM INSTALLATION ERRORS Sometimes the installation of RPM software doesn’t go according to plan and you need to take corrective actions. Failed Dependencies Sometimes RPM installations will fail. This section shows you how to recover from some of the most common errors you’ll encounter. In the next example we install the MySQL client ignoring dependencies.23.rpm error: Failed dependencies: libmysqlclient.58-9. To get around this problem you can run the rpm command with the --nodeps option to disable dependency checks.9003-4 [root@bigboy tmp]# Strangely enough.58-9. which fails because the mysql MySQL client RPM. ####################### [100%] 1:mysql ####################### [100%] .23..rpm error: Failed dependencies: perl-DBD-MySQL is needed by mysql-3. giving Failed dependencies errors..10 is needed by perl-DBD-MySQL-2.23. the installation of the perl-DBD-MySQL package fails because it needs the mysql client package. You will then have to install the compiled RPMs found in their respective subdirectories as you normally would.rpm error: Failed dependencies: libmysqlclient.rpm Preparing.23. on which it depends.i386.23. followed by successful installation of perl-DBD-MySQL and mysql-server: [root@bigboy tmp]# rpm -Uvh --nodeps mysql-3.9003-4. In the next example we’re attempting to install the MySQL database server application.58-9 mysql = 3.58-9 [root@bigboy tmp]# rpm -Uvh perl-DBD-MySQL-2.58-9. For example.23. if you compiled an i386 architecture version of the RPM it will be placed in the i386 subdirectory.so.58-9 [root@bigboy tmp]# MySQL Installing the MySQL client also fails because it requires the package: perl-DBD- [root@bigboy tmp]# rpm -Uvh mysql-3.qxd 2/25/05 9:58 AM Page 115 RPM Installation Errors 115 The compiled RPM file can now be found in one of the architecture subdirectories under /usr/src/redhat/RPMS directory.so.i386.58 is needed by mysql-server-3.i386.i386. needs to be installed beforehand: [root@bigboy tmp]# rpm -Uvh mysql-server-3.10 is needed by mysql-server-3.

Signature Keys Fedora digitally signs all its RPM files.i386.16. One of the advantages of using the yum facility is that you don’t have to worry about this dependency process as much because the dependency RPMs are always downloaded and installed automatically also.0pl2-6.i386.07Harrison_ch06. key ID 4f2a6fd2 error: dhcp-3.rpm cannot be installed [root@bigboy tmp]# ... #################################### [100%] 1:dhcp #################################### [100%] [root@bigboy tmp]# It is always good to install the key files.23.qxd 2/25/05 9:58 AM Page 116 116 Installing RPM Software Chapter 6 [root@bigboy tmp]# rpm -Uvh perl-DBD-MySQL-2.rpm error: dhcp-3..0pl2-6. so it’s best to import their public encryption key beforehand so that the RPM installation program will be able to verify the validity of the RPM file. key ID 4f2a6fd2 Preparing. ####################### [100%] 1:perl-DBD-MySQL ####################### [100%] [root@bigboy tmp]# rpm -Uvh mysql-server-3. If the RPM’s digital signature doesn’t match that in the key file.i386.rpm Preparing. the rpm command can automatically install all the prerequisite RPMs using the --aid option.i386.. the rpm installation program also alerts you and fails to install the RPM package at all: [root@bigboy tmp]# rpm -Uvh dhcp-3..0pl2-6..0pl2-6.rpm warning: dhcp-3. If they are not there.rpm: V3 DSA signature: NOKEY.i386. ####################### [100%] 1:mysql-server ####################### [100%] [root@bigboy tmp]# Note If all the installation RPMs are located in the same directory. It is a good idea to import both the Red Hat and Fedora keys: [root@bigboy tmp]# rpm --import /usr/share/rhn/RPM-GPG-KEY [root@bigboy tmp]# rpm --import /usr/share/rhn/RPM-GPG-KEY-fedora [root@bigboy tmp]# If you don’t install the keys you get a DSA signature warning that alerts you to the fact that the RPM file might be bogus: [root@bigboy tmp]# rpm -Uvh dhcp-3.i386.16.i386.58-9. This can be done using the rpm command as seen in the next example.16.16.rpm: V3 DSA signature: BAD.rpm Preparing.16.9003-4. the RPMs will install with only a warning message.0pl2-6.

. .4p1-2 openssh-3.2.07Harrison_ch06. In this example we are looking for all packages containing the string ssh in the name.. [root@bigboy tmp]# rpm -qa | grep -i ssh openssh-server-3. It is therefore important to be able to determine the origin of certain files by listing the contents for RPMs in which you suspect the files might reside. If you are not sure of the package name and its capitalization. HOW TO LIST INSTALLED RPMS The rpm -qa command lists all the packages installed on your system: [root@bigboy tmp]# rpm -qa perl-Storable-1.27-9 libstdc++-3.4p1-2 openssh-clients-3.3-3 . the latter method is probably more suitable. [root@bigboy tmp]# You can also pipe the output of this command through the grep command if you are interested in only a specific package.14-15 smpeg-gtv-0.4p1-2 [root@bigboy tmp]# Note You could use the rpm -q package-name command to find an installed package because it is much faster than using grep and the -qa switch. .4p1-2 openssh-askpass-gnome-3. LISTING FILES ASSOCIATED WITH RPMS Sometimes you’ll find yourself installing software that terminates with an error requesting the presence of a particular file..4..4p1-2 openssh-askpass-3. . but you have to have an exact package match.0.qxd 2/25/05 9:58 AM Page 117 Listing Files Associated with RPMs 117 Signatures are therefore useful because they help protect you against tampered and otherwise corrupted RPMs being installed..4-9 e2fsprogs-1.. regardless of case (-i means ignore case).2-7 audiofile-0. In many cases the installation program doesn’t state the RPM package in which the file can be found.

.0pl1-23.2-0. You can use the -qpl qualifier to list all the files in an RPM archive to make sure before installing it: [root@bigboy updates]# rpm -qpl dhcp-3.1.. and then we use the -ql qualifier to get the file listing: [root@bigboy tmp]# rpm -q ntp ntp-4.htm [root@bigboy tmp]# Listing Files in RPM Files Sometimes you make a guess and download what you think is the RPM with the missing file.rpm /etc/rc.i386.2 [root@bigboy tmp]# rpm -ql ntp /etc/ntp /etc/ntp. .gz /var/lib/dhcp /var/lib/dhcp/dhcpd.07Harrison_ch06... . /usr/share/man/man8/dhcrelay. .htm /usr/share/doc/ntp-4. In this example we test to make sure that the NTP package is installed using the-q qualifier. .8. /usr/share/doc/ntp-4.1.2/rdebug.2/release.. In this case you need to know which RPM on the old server contains the file...htm /usr/share/doc/ntp-4.d/init..htm /usr/share/doc/ntp-4..rc1..d/dhcrelay /etc/sysconfig/dhcpd /etc/sysconfig/dhcrelay .qxd 2/25/05 9:58 AM Page 118 118 Installing RPM Software Chapter 6 Listing Files for Already Installed RPMs This can be useful if you have to duplicate a working server that is already in a production environment.2/refclock..2/tickadj.d/dhcpd /etc/rc.1.1.leases [root@bigboy updates]# . Sometimes the installation of an application fails on the new server due to the lack of a file that resides on the old one.conf /etc/ntp/drift /etc/ntp/keys .d/init.1. You can use the -ql qualifier to list all the files associated with an installed RPM..

the command to use is tar -xzvf filename. their file extensions end with .tar.tgz.cnf mysql-3. which can frequently be obtained from the Web sites of software developers.qxd 2/25/05 9:58 AM Page 119 Installing Software Using tar Files 119 Listing the RPM to Which a File Belongs You might need to know the RPM that was used to install a particular file. The commands to extract the data from either type are similar. A more detailed explanation will be provided in Chapter 7. The package name given must match that listed in the rpm -qa command because the version of the package is important: [root@bigboy tmp]# rpm -e package-name WHICH RPMS START AT BOOT TIME? The best way to view and configure which RPMs will start at boot time is by using the chkconfig command with the --list switch.tar. When a tar file is uncompressed. and online software libraries such as www.23. not a file named /etc/mysql.” which covers the Linux boot process.sourceforge. The following example confirms the origin of the /etc/my. This is useful when you have a suspicion about the function of a file but are not entirely sure. The Linux tar command is used to archive files. When the archive is compressed.tar file extension in the filename.58-9 [root@zippy tmp]# UNINSTALLING RPMS The rpm -e command erases an installed package.gz or . INSTALLING SOFTWARE USING TAR FILES Another popular software installation file format is the tar file.conf as you’d normally expect. the command to extract the data is tar -xvf filename. For example. which typically have a . “The Linux Boot Process. the MySQL RPM uses the /etc/my.07Harrison_ch06. These files are also frequently compressed in the gzip format.net. .tar.cnf file as its configuration file.gz.cnf file: [root@zippy tmp]# rpm -qf /etc/my. and when they are.

configuration. mail server. .pm Makefile.tar.3. which outlines all the customized steps to install the software.in Software installation with tar files can be frustrating.qxd 2/25/05 9:58 AM Page 120 120 Installing RPM Software Chapter 6 The tar file installation process usually requires you first to uncompress and extract the contents of the archive in a local subdirectory..in NEWS [root@bigboy linux-software-1.1 linux-software-1. CONCLUSION This is just the beginning. linux-software-1. RPMs. Issue the tar command to extract the files: [root@bigboy tmp]# tar -xvzf linux-software-1.07Harrison_ch06. If the software you install is intended to make your Linux machine permanently run an application—such as a Web server.1/linux-software-plugins.spec Helper. Here are the initial steps to install tar-based software: 1.3..3. Use the cd command to enter the subdirectory and follow the directions listed in the INSTALL and README files: [root@bigboy tmp]# cd linux-software-1. and troubleshooting of many of the most popular Linux server applications used today. with the single standardized command format.3. or any other type of server—you have to know how to get the software activated when the system reboots. testing. Subsequent chapters cover the use.1..3.spec. each with its own customized installation commands..1. .tar...in INSTALL Makefile.am linux-software. .3.gz [root@bigboy tmp]# 2. frequently requiring the installation of other supporting tar files.3.1/ linux-software-1.1]# ls COPYING install-sh missing depcomp LEGAL mkinstalldirs scripts FAQ lib linux-software. This is covered in Chapter 7.spec [root@bigboy tmp]# This creates a subdirectory with the installation files inside: [root@bigboy tmp]# ls linux-software-1. The subdirectory usually contains a file called README or INSTALL.gz linux-software-1. are usually easier to use and may be the better method to use for newer Linux users.3.1]# plugins pluginsREADME REQUIREMENTS subst.3. which frequently has the same name as the tar file.1 [root@bigboy linux-software-1.1/plugins-scripts/ .

for each partition is always reserved for programmable code used in booting. the PC system BIOS loads and executes the boot loader code in the MBR. When booting from a hard disk. THE FEDORA BOOT SEQUENCE You might remember when you installed Linux that the installation process prompted you for a list of partitions and the sizes of each in which your filesystems would be placed. Fedora Linux is supplied with the GRUB boot loader. When allocating disk space for the partitions. the first sector. 121 . When you have this information you can use it to alter the type of login screen you get as well as which programs start up. The very first sector of the hard disk is reserved for the same purpose and is called the master boot record (MBR). or data unit.08Harrison_ch07. Read on for the details. which is fairly sophisticated and therefore cannot entirely fit in the 512 bytes of the MBR.qxd 2/25/05 9:58 AM Page 121 C H A P T E R 7 The Linux Boot Process IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ The Fedora Boot Sequence Determining the Default Boot runlevel Getting a GUI Console Get a Basic Text Terminal Without Exiting the GUI System Shutdown and Rebooting How to Set Programs to Run at Each runlevel Conclusion Learning how Linux boots up is critical. The MBR then needs to know which partitions on the disk have boot loader code specific to their operating systems in their boot sectors and then attempts to boot one of them.

1 Example of a grub.d /etc/rc. it first runs the /sbin/init program.0)/grub/splash.521 ro root=LABEL=/ initrd /initrd-2.521) root (hd0.8-1.8-1.d/rc6.conf file for a system that can boot both Fedora Linux and Windows 2000.1.6. When Linux begins to boot with its kernel. This then reads the data in the /boot/grub/grub.1 shows a typical grub.qxd 2/25/05 9:58 AM Page 122 122 The Linux Boot Process Chapter 7 The GRUB MBR boot loader merely searches for a special boot partition and loads a second stage boot loader. The structure of this file is discussed further in Chapter 33. It then inspects the /etc/inittab file to determine Linux’s overall mode of operation or runlevel.d /etc/rc. the second stage boot loader then displays the familiar Fedora branded splash screen that lists all the configured operating system kernels for your choice.img title Windows 2000 rootnoverify (hd0.conf file. “Modifying the Kernel to Improve Performance. and starts vital programs needed for the operating system to function properly. Figure 7.d /etc/rc.d/rc2.08Harrison_ch07.d/rc0.d /etc/rc.d/rc3. A listing of valid runlevels can be seen in Table 7.0) kernel /vmlinuz-2. which lists all the available operating systems and their booting parameters.1) chainloader +1 Figure 7.xpm.d/rc5.conf configuration file.d/rc1.gz title Fedora Core (2. Table 7. which does some system checks such as verifying the integrity of the file systems.1 Linux Runlevels Directory /etc/rc.d /etc/rc.8-1.6. When this is complete.d /etc/rc.521.” default=0 timeout=10 splashimage=(hd0.d Mode/Runlevel 0 1 2 3 4 5 6 Runlevel Description Halt Single-user mode Not used (user-definable) Full multiuser mode (no GUI interface) Not used (user-definable) Full multiuser mode (with GUI interface) Reboot .d/rc4.6.

. .qxd 2/25/05 9:58 AM Page 123 Determining the Default Boot runlevel 123 Based on the selected runlevel. .. K95firstboot S17keytable [root@bigboy tmp]# ..reboot (Do NOT set initdefault to this) . Most Red Hat/Fedora packages place their startup script in the /etc/init. Here is a directory listing of the scripts in the /etc/rc. when set to 5.... This is explained later. The number that follows the K or S specifies the position in which the scripts should be run. . you get the GUI..d/rc6.. DETERMINING THE DEFAULT BOOT RUNLEVEL The default boot runlevel is set in the file /etc/inittab with the initdefault variable... it won’t be run. .Multiuser. .d . the init process then executes startup scripts located in subdirectories of the /etc/rc...Full multiuser mode # 4 .. .. ... if you do not have networking) # 3 ..d directory: [root@bigboy tmp]# ls /etc/rc. . in ascending order. which means the script should be run when the system is shutting down.. . each filename in these directories either starts with an S.d.. This makes file management a lot easier. K91isdn S10network . The runlevels used by RHS are: # 0 . . .. . which can then be used for another day. The deletion of a link doesn’t delete the file. without NFS (The same as 3. ..Single user mode # 2 . As you can see. In our example... K92iptables S12syslog .d/rc0. K75netfs K96pcmcia ..d...... . ..d through /etc/rc.... When set to 3. K87portmap S09wlan .d/rc3... or a K.....unused # 5 . kudzu with a value of 05 will run before wlan with a value of 09.. K86nfslock S05kudzu . the system boots up with the text interface on the VGA console. .08Harrison_ch07. Scripts used for runlevels 0 to 6 are located in subdirectories /etc/rc.X11 # 6 .d directory... . Here is a snippet of the file (delete the initdefault line you don’t need): # Default runlevel.... If a script isn’t there.d directory and place symbolic links (pointers) to this script in the appropriate subdirectory of /etc/rc.. respectively.halt (Do NOT set initdefault to this) # 1 . which signifies the script should be run at startup... Fortunately you don’t have to be a scripting/symbolic linking guru to make sure everything works right because Fedora comes with a nifty utility called chkconfig to do it for you.d/rc3.

☞ Of course. or servers that might have to be operated for an extended period of time by relatively nontechnical staff. ☞ Changing initdefault from 3 to 5. Remember that when you log out you will get the regular text-based console again: [root@bigboy tmp]# startx ☞ Automatic Method: You can have Linux automatically start the X terminal GUI console for every login attempt until your next reboot by using the init command.08Harrison_ch07. using the GUI only as necessary with the startx command. may need to be run at runlevel 5 all the time through the init 5 command. you might want to operate in text mode runlevel 3 most of the time. has an effect on your next reboot. Setting it to 0 will never allow it to start! GETTING A GUI CONSOLE You have two main options if your system comes up in a text terminal mode on the VGA console and you want to get the GUI: ☞ Manual Method: You can start the X terminal GUI application each time you need it by running the startx command at the VGA console. . or vice versa. to keep this functionality even after you reboot.qxd 2/25/05 9:58 AM Page 124 124 The Linux Boot Process Chapter 7 # id:3:initdefault: id:5:initdefault: # Console Text Mode # Console GUI Mode Note the following: ☞ Most home users boot up with a Windows like GUI (runlevel 5). Remember you can make runlevel 5 permanent even after a reboot by editing the /etc/inittab file. You will need to edit your initdefault variable in your /etc/inittab file. See the following section on how to get a GUI login all the time until the next reboot. as mentioned in the preceding section. Servers that double as personal workstations. don’t set the initdefault value to 6 or your system will constantly reboot. ☞ Most techies tend to boot up with a plain text-based command-line-type interface (runlevel 3). [root@bigboy tmp]# init 5 When the CPU capacity or available memory on your server is low or you want to maximize all system resources.

or worse. This makes the GUI run as number 7: ☞ You can step through each virtual console session by using the Ctrl-AltF1 through F6 key sequence. SYSTEM SHUTDOWN AND REBOOTING It is usually not a good idea to immediately power off your system when you are finished using it. Halt/Shut Down the System The init command enables you to change the current runlevel. This can be important if you need quick access to commands or you are not familiar with the GUI menu option layout. ☞ Click on Systems Tools and then Terminal. The X terminal GUI console creates its own virtual console using the first available TTY that is not controlled by mingetty. You’ll get a new login prompt for each attempt. which will be outlined in this section. Using a GUI Terminal Window You can open a GUI-based window with a command prompt inside by doing the following: ☞ Click on the Fedora logo button in the lower-left corner of the screen. Linux has a number of ways to gracefully shut down and reboot your system.08Harrison_ch07. you could corrupt the filesystem directory structure. or if the GUI is running after launching startx. that value is 0. ☞ You can get the GUI login with the sequence Ctl-Alt-F7. Using Virtual Consoles By default. only in runlevel 5. and for a shutdown.qxd 2/25/05 9:58 AM Page 125 System Shutdown and Rebooting 125 GET A BASIC TEXT TERMINAL WITHOUT EXITING THE GUI There are a number of ways for you to get a command prompt when running a Linux GUI. This can cause files that are being updated to become corrupted. These are defined by the mingetty statements in the /etc/inittab file. Linux runs six virtual console or TTY sessions running on the VGA console. Here is an example: [root@bigboy tmp]# init 0 .

which can be used to the same effect.08Harrison_ch07. please refer to the man pages for details. but it also sends a warning message to all users: [root@bigboy tmp]# reboot Broadcast message from root (pts/0) (Sat Nov The system is going down for reboot NOW! [root@bigboy tmp]# 6 12:39:31 2004): -r More graceful reboots can be done with the shutdown command using the switch and specifying a delay. which in this case is 10 minutes: [root@bigboy root]# shutdown -ry 10 Broadcast message from root (pts/0) (Sat Nov 6 13:26:39 2004): The system is going DOWN for reboot in 10 minutes! Broadcast message from root (pts/0) (Sat Nov 6 13:27:39 2004): The system is going DOWN for reboot in 9 minutes! . Another advantage of the shutdown command is that it warns people that the shutdown is going to occur: [root@bigboy tmp]# shutdown -hy 0 Broadcast message from root (pts/0) (Sat Nov The system is going down for system halt NOW! [root@bigboy tmp]# 6 13:15:27 2004): Reboot the System You can also use the entering runlevel 6: init command to reboot the system immediately by [root@bigboy tmp]# init 6 The reboot command has the same effect. You can also specify shutting down at a specific time of the day. which can be avoided with the -y switch. It often prompts you as to whether you are sure you want to execute the command. and the first argument tells it how long to wait before starting the procedure. The -h switch forces the system to halt.qxd 2/25/05 9:58 AM Page 126 126 The Linux Boot Process Chapter 7 Fedora also has a shutdown command. in this case 0 minutes.

. ...05b# .. bash-2.. Broadcast message from root (pts/0) (Sat Nov The system is going down for reboot NOW! 6 13:36:39 2004): Entering Single-user Mode Some activities require you to force the system to log off all users. third-party applications.qxd 2/25/05 9:58 AM Page 127 System Shutdown and Rebooting 127 . Another reason is the recovery of your root password. bash-2. .. as mentioned in Chapter 27. Switching to Single-user Mode When the system is running normally.... and the shutdown command doesn’t have a single-user mode option. ..08Harrison_ch07. this gives no prior warning to users... A typical scenario is the addition of a new hard disk. This can be overcome by running the shutdown command with a delay in minutes as the only argument: [root@bigboy tmp]# shutdown 1 Broadcast message from root (pts/0) (Sat Nov 6 13:44:59 2004): The system is going DOWN to maintenance mode in 1 minute! Broadcast message from root (pts/0) (Sat Nov 6 13:45:59 2004): The system is going down to maintenance mode NOW! . this can be done by using the init command to enter runlevel 1.05b# Unfortunately.. and networking so that only the systems administrator has access to the system from the VGA console.” or the troubleshooting of a failed boot process... “Expanding Disk Capacity. because if you do it from a remote terminal session you’ll be logged out: [root@bigboy root]# init 1 . It is best to do this from the console.

or the previous systems administrator might move on to a new job without giving it to you. init 3 and init 5) to alter this default behavior: bash-2. .. The system will then shut down in an orderly fashion. Fedora Core release 2 (Tettnang) Kernel 2..521 on an i686 bigboy login: Root Password Recovery Sometimes you might forget the root password.. Reverting to Your Default runlevel from Single-user Mode The exit command forces the system to exit runlevel 1 and revert to the default runlevel for the system. change your password. The system will continue to boot. When at the command prompt.8-1.. so you don’t have to specify a root username.” You will get a prompt like this to add boot options to the boot string found in your grub.qxd 2/25/05 9:58 AM Page 128 128 The Linux Boot Process Chapter 7 Entering Single-user Mode at the Grub Splash Screen You can enter single-user mode directly after turning on the power to your system: 1. follow these steps: 1. Press enter. Reboot the system and enter single-user mode. 3. Single-user mode assumes the person at the console is the systems administrator root. 2. Power on your system.05b# exit INIT: Entering runlevel: 3 . but will go straight to the root # prompt without first asking for a username and password. Return to your default runlevel by using the exit command.6. Go to the VGA console and press Ctrl-Alt-Del. 4.conf file: grub append> ro root=LABEL=/ 3. ... You can also use the init command (for example.08Harrison_ch07. the one which allows you to select the various installed kernels on the system. type the letter a for “append. . When the Linux logo screen appears. 2. Add a 1 to the end of the string: grub append> ro root=LABEL=/ 1 4. To recover the root password.

Use chkconfig to Get a Listing of sendmail’s Current Startup Options The chkconfig command can be used with grep to determine the runlevels in which sendmail will run. the chkconfig command can be used to adjust which applications start at each runlevel.d and the runlevels at which they will be on or off: [root@bigboy tmp]# chkconfig --list keytable 0:off 1:on 2:on 3:on 4:on atd 0:off 1:off 2:off 3:on 4:on syslog 0:off 1:off 2:on 3:on 4:on gpm 0:off 1:off 2:on 3:on 4:on kudzu 0:off 1:off 2:off 3:on 4:on wlan 0:off 1:off 2:on 3:on 4:on sendmail 0:off 1:off 2:off 3:on 4:off netfs 0:off 1:off 2:off 3:on 4:on network 0:off 1:off 2:on 3:on 4:on random 0:off 1:off 2:on 3:on 4:on .. and the second defines whether you want it on or off..08Harrison_ch07..qxd 2/25/05 9:58 AM Page 129 How to Set Programs to Run at Each runlevel 129 HOW TO SET PROGRAMS TO RUN AT EACH RUNLEVEL As stated earlier. Here we see it will run at levels 3 and 5: [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:on 4:off 5:on 6:off [root@bigboy tmp]# Switch Off sendmail Starting Up in Levels 3 and 5 The chkconfig command with the --level switch indicates that some action needs to be done at the runlevels entered as its values. . 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 5:on 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off 6:off chkconfig Examples You can use chkconfig to change runlevels for particular packages. Let’s change it so that sendmail doesn’t start up at boot. Here we see that sendmail starts with a regular startup at runlevel 3 or 5. In this case we want sendmail not to be started when entering runlevels 3 and 5: [root@bigboy tmp]# chkconfig --level 35 sendmail off [root@bigboy tmp]# . The first argument in the command is the package you want to affect.. You can use this command with the --list switch to get a full listing of packages listed in /etc/init.

08Harrison_ch07. we can use chkconfig once more. makes the changes for runlevels 3 and 5 automatically: [root@bigboy tmp]# chkconfig sendmail off chckconfig Because the intention is to shut down also have to stop it from running now: sendmail permanently. This usually results in your system listening on a variety of unexpected TCP/IP ports that could be used as doors into your system by hackers. Start sendmail again to get it running immediately. The screen output of the netstat -an command shows a typical case.qxd 2/25/05 9:58 AM Page 130 130 The Linux Boot Process Chapter 7 By not specifying the runlevels with the --level switch. we might [root@bigboy tmp]# service sendmail stop Shutting down sendmail: [ OK ] Shutting down sm-client: [ OK ] [root@bigboy tmp]# Double-check that sendmail Will Not Start Up We can then use chkconfig to double-check our work: [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:off 4:off 5:off 6:off [root@bigboy tmp]# Turn On sendmail Again To reactivate sendmail. but some others are less obvious. [root@bigboy tmp]# chkconfig sendmail on [root@bigboy tmp]# chkconfig --list | grep mail sendmail 0:off 1:off 2:off 3:on 4:off 5:on 6:off [root@bigboy tmp]# service sendmail start Starting sendmail: [ OK ] Starting sm-client: [ OK ] [root@bigboy tmp]# Using chkconfig to Improve Security A default Fedora installation automatically starts a number of daemons that you may not necessarily need for a Web server. respectively. TCP ports 25 and 22 are for mail and SSH. not just after the next reboot. Some ports are relatively easy to recognize. but with the on argument. Should you use the chkconfig command and the scripts in the /etc/init.d directory to shut these down permanently? .

and ntp daemons.0..08Harrison_ch07.0:* LISTEN tcp 0 0 0.0.0.. [root@bigboy root]# .0.0:* udp 0 0 0.0.0.0.0:* udp 0 0 0.1:32769 0. “Remote Disk Access with NFS.qxd 2/25/05 9:58 AM Page 131 How to Set Programs to Run at Each runlevel 131 [root@bigboy tmp]# netstat -an Active Internet connections (servers and established) Proto Recv-Q Send-Q Local Address Foreign Address State tcp 0 0 0.0:* LISTEN tcp 0 0 127. topics that are covered in Chapters 29.1:25 0.0.0.0.0:* .0:32768 0.0.0:* LISTEN tcp 0 0 127.0.0.0:* udp 0 0 0.. In the next examples.0:32768 0.0.0.0. “The NTP Server.0. If you don’t need any of these three applications. “Configuring NIS.0.0. and XINETD.” and it might be required for their operation: [root@ bigboy tmp]# lsof -i tcp:111 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME portmap 1165 rpc 4u IPv4 2979 TCP *:sunrpc (LISTEN) [root@ bigboy tmp # [root@bigboy COMMAND PID xinetd 1522 (LISTEN) [root@bigboy tmp]# lsof -i tcp:32769 USER FD TYPE DEVICE SIZE NODE NAME root 5u IPv4 2764 TCP probe-001:32769 tmp]# [root@bigboy root]# lsof -i udp:123 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME ntpd 1321 ntp 4u IPv4 3390 UDP *:ntp .” and 30.0:* LISTEN tcp 0 0 :::22 :::* LISTEN udp 0 0 0.0.0. which lists all open or actively used files and can be given additional options to extend its scope to include the TCP/IP protocol stack.0. which is covered in Chapter 24. “Telnet..0:* LISTEN tcp 0 0 127. it’s best to shut down portmap permanently.. we see that TCP ports 111 and 32769.” portmap also has many known security flaws that makes it advisable to be run on a secured network. The portmap daemon is required for the operation of NFS and NIS. and UDP port 123 are being used by the portmap.0.0:68 0.0:111 0.0. . [root@bigboy tmp]# For example..0..0.0.0.0.0. A number of network applications are reliant on xinetd.0. TFTP.0:631 0.1:631 0.0.0. .0. ntp.0:930 0..0:111 0.0.0.0:* udp 0 0 0. how do you know which startup script is responsible for TCP port 111? The answer is to use the lsof command.0.0.0. xinetd.0. as explained in Chapter 16.” is required for synchronizing your time with a reliable time source and may be necessary. respectively.

we’ve discovered that TCP port 32768 is being used by rpc. [root@bigboy tmp]# As a rule of thumb. ☞ Don’t add/remove anything to other runlevels unless you absolutely know what you are doing. I usually shut down nfs. We soon discover that the nfslock daemon uses it.0.0.d directory. depicted as IP address 0. . In the example below.statd ls: /etc/init. but you should always have a security reference guide on hand to guide your final decisions.d/nfslock:[ -x /sbin/rpc. applications listening only on the loopback interface (IP address 127. and patch updates need to be considered.0. Remember to research your options thoroughly before choosing to shut down an application. and the Internet for information. are naturally more vulnerable and their continued operation should be dependent on your server’s needs. Those listening on all interfaces.statd. reference books. “Configuring Linux Mail Servers. portmap.. . password policies. Your needs may be different. I keep sendmail running as it is always needed to send and receive mail (see Chapter 21.1) are usually the least susceptible to network attack and probably don’t need to be stopped for network security reasons.qxd 2/25/05 9:58 AM Page 132 132 The Linux Boot Process Chapter 7 In some cases it’s tricky to determine the application based on the results of the lsof command. Final Tips on chkconfig Remember the following: ☞ In most cases. unless in a test environment.d/rpc. physical access restrictions. netfs. you want to modify runlevels 3 and 5 simultaneously and with the same values.08Harrison_ch07. If you don’t need nfslock.0.. Firewalls.” for details).d/rpc. Unpredictable results are always undesirable.statd file in the /etc/init. Shutting down applications is only a part of server security.statd: No such file or directory [root@bigboy tmp]# grep -i statd /etc/init. Don’t experiment. and cups printing as standard practice on Internet servers.d/* /etc/init.statd 1178 rpcuser 6u IPv4 2400 TCP *:32768 (LISTEN) [root@bigboy tmp]# ls /etc/init. nfslock.statd ] || exit 0 .. Use the Linux man pages. The simple solution is to use the grep command to search all the files for the string rpc.0.statd to determine which one is responsible for its operation. Full coverage of server and network security is beyond the scope of this book. but there is no rpc.. then shut it down permanently: [root@bigboy tmp]# lsof -i tcp:32768 COMMAND PID USER FD TYPE DEVICE SIZE NODE NAME rpc.

which was covered in Chapter 5. The commands for starting and stopping the programs covered in this book are covered in each respective chapter.qxd 2/25/05 9:58 AM Page 133 Conclusion 133 ☞ chkconfig doesn’t start the programs in the /etc/init. but like syslog. Sometimes they start but give errors listed only in the /var/log directory. Whenever possible. CONCLUSION The topics discussed in this chapter might seem simple.d directory.” they are an essential part of Linux administration that gets frequently overlooked— especially when new software is installed. . “Troubleshooting Linux with syslog. Taking the time to configure and test your startup scripts could prevent you from being awakened in the middle of the night while you are on vacation! It is really important.08Harrison_ch07. it just configures them to be started or ignored when the system boots up. always try to reboot your system to make sure all the newly installed applications start up correctly.

qxd 2/25/05 9:58 AM Page 134 .08Harrison_ch07.

If you install a home cable/DSL router between your modem and home network. You can choose to disable the DHCP server feature on your home router and set up a Linux box as the DHCP server. Chapter 6.” covers how to do this in detail.qxd 2/25/05 9:59 AM Page 135 C H A P T E R 8 Configuring the DHCP Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install the DHCP Pacakge The /etc/dhcpd.09Harrison_ch08. If you need a refresher. Downloading and installing RPMs isn’t hard. The configuration of a Linux DHCP client that gets its IP address from a DHCP server is covered in Chapter 3.conf File How to Get DHCP Started DHCP Servers with Multiple NICs Configuring Linux Clients to Use DHCP Configuring Windows Clients to Use DHCP Simple DHCP Troubleshooting Conclusion Normally if you have a cable modem or DSL. you get your home PC’s IP address dynamically assigned from your service provider. 135 . This chapter covers only the configuration of a DHCP server that provides IP addresses. your PC will most likely get its IP address at boot time from the home router instead. “Linux Networking.” DOWNLOAD AND INSTALL THE DHCP PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format. “Installing RPM Software.

conf in the following directory. Here is the command to do the copying for the version 3.201 192. but you can find a sample copy of dhcpd.168.qxd 2/25/05 9:59 AM Page 136 136 Configuring the DHCP Server Chapter 8 When searching for the file.0p11 RPM file: [root@bigboy tmp]# cp /usr/share/doc/dhcp-3.conf file to the /etc directory and then you have to edit it.0pl1/dhcpd. remember that the DHCP server RPM’s filename usually starts with the word dhcp followed by a version number like this: dhcp-3. The standard DHCP RPM package doesn’t automatically install a /etc/dhcpd.1.1. It uses the commands here to configure your network.conf file: Most importantly. # Don’t forward DHCP requests from this # NIC interface to any other NIC # interfaces option ip-forwarding off.conf Here is a quick explanation of the dhcpd.1rc14-1. which you can always use as a guide: /usr/share/doc/dhcp-<version-number>/dhcpd.168.conf.sample You have to copy the sample dhcpd.conf. it reads the file /etc/dhcpd. # Set the default gateway to be used by # the PC clients option routers 192.255.1.0.rpm. # Set the amount of time in seconds that # a client may keep the IP address default-lease-time 86400.09Harrison_ch08.168.168.i386.220.conf. . THE /ETC/DHCPD.1.conf file. there must be a subnet section for each interface on your Linux box: ddns-update-style interim ignore client-updates subnet 192.CONF FILE When DHCP starts. max-lease-time 86400.0 netmask 255.sample \ /etc/dhcpd.255.1.0 { # The range of IP addresses the server # will issue to DHCP enabled PC clients # booting up on the network range 192.

fixed-address 192. } } # # List an unused interface here # subnet 192. These include telling the DHCP clients where to go for services such as finger and IRC. Some devices such as network printers default to getting their IP addresses using DHCP.1.100.1.255. option subnet-mask 255.0 netmask 255.qxd 2/25/05 9:59 AM Page 137 The /etc/dhcpd. Check the dhcp-options man page after you do your install: [root@bigboy tmp]# man dhcp-options Note The host statement seen in the sample dhcpd. # You can also assign specific IP addresses based on the clients’ # ethernet MAC address as follows (Host’s name is “laser-printer”: host laser-printer { hardware ethernet 08:00:2b:4c:59:23.0.conf file can be very useful.168.168. # you need to include the following option in the dhcpd.255.222. This statement can be used to always provide specific IP address to DHCP queries from a predefined NIC MAC address.1.255. but users need to access them by a fixed IP address to print their documents. .168.100.100.255.conf file: option netbios-name-servers 192.0 { } There are many more options statements you can use to configure DHCP. # If you specify a WINS server for your Windows clients.168.conf File 137 # Set the broadcast address and subnet mask # to be used by the DHCP clients option broadcast-address 192.1.1.2.168.168. # Set the DNS server to be used by the # DHCP clients option domain-name-servers 192. This can help to reduce systems administration overhead. # Set the NTP server to be used by the # DHCP clients option nntp-server 192.09Harrison_ch08.255.

255.255.255 address so that it knows the interface on which to send the reply. Note You can’t run your DHCP sever on multiple interfaces because you can have only one route to network 255. If you try to do it.qxd 2/25/05 9:59 AM Page 138 138 Configuring the DHCP Server Chapter 8 HOW TO GET DHCP STARTED To start DHCP: 1. DHCP SERVERS WITH MULTIPLE NICS When a DHCP configured PC boots. (In both of the next two examples.leases to create the file if it does not exist: [root@bigboy tmp]# touch /var/lib/dhcp/dhcpd.255.255. Use the command touch /var/lib/dhcp/dhcpd. You also can test whether the DHCP process is running with the following command. It does this by sending a standardized DHCP broadcast request packet to the DHCP server with a source IP address of 255. you have to add a route for this 255.255. always remember to set your PC to get its IP address via DHCP. it sends it to the default gateway. it requests its IP address from the DHCP server. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep dhcpd 5.255.leases file. Remember to restart the DHCP process every time you make a change to the conf file for the changes to take effect on the running process.09Harrison_ch08. we assume that DHCP requests will be coming in on interface eth0.255. Some older Fedora/Red Hat versions of the DHCP server will fail unless there is an existing dhcpd. . Finally.leases 2. if not. you’ll discover that DHCP server working on only one interface. Use the chkconfig command to get DHCP configured to start at boot: [root@bigboy tmp]# chkconfig dhcpd on 3. Use the service command to instruct the start/stop/restart DHCP after booting: [root@bigboy tmp]# service dhcpd start [root@bigboy tmp]# service dhcpd stop [root@bigboy tmp]# service dhcpd restart /etc/init. If your DHCP server has more than one interface.d/dhcpd script to 4.255.) Note More information on adding Linux routes and routing may be found in Chapter 3.

CONFIGURING WINDOWS CLIENTS TO USE DHCP Fortunately. use the permanent solution that’s discussed next.255. Permanent Solution The new Fedora Linux method of adding static routes doesn’t seem to support sending traffic out an interface that’s not destined for a specific gateway IP address.255 using the command: [root@bigboy tmp]# route add -host 255.255. The following sections present simple troubleshooting steps that you can go through to ensure that DHCP is working correctly on your network. Please refer to this chapter if you need a quick refresher on how to configure a Linux DHCP client.09Harrison_ch08. To avoid this problem add the route add command to your /etc/rc. CONFIGURING LINUX CLIENTS TO USE DHCP A Linux NIC interface can be configured to obtain its IP address using DHCP with the examples outlined in Chapter 3.255.255. Windows defaults to using DHCP for all its NIC cards so you don’t have to worry about doing any reconfiguration.255.qxd 2/25/05 9:59 AM Page 139 Simple DHCP Troubleshooting 139 Temporary Solution You can temporarily add a route to 255.255 isn’t intended to be relayed to a gateway.local script. after the server is configured correctly there is no need to change any settings and it therefore runs reliably.255. The DHCP packet destined for address 255. SIMPLE DHCP TROUBLESHOOTING The most common problems with DHCP usually aren’t related to the server. The problems usually occur at the DHCP client’s end for a variety of reasons.255 dev eth0 route add If you want this routing state to be maintained after a reboot. . but it should be sent using the MAC address of the DHCP client in the Ethernet frame.

If you cannot. but it is an interesting project to try.255. but before we do. Just remember to make sure that the range of IP addresses issued by all DHCP servers on a network doesn’t overlap because it could possibly cause unexpected errors. 2. we’ll do a quick refresher on how to create the Linux users who’ll be using many of the applications outlined in the rest of the book. Ensure that your DHCP server is configured correctly and use the pgrep command discussed earlier to make sure the DHCP process is running. Now it’s time to try something harder.255. Creating a Linux DHCP server is straightforward and touches all the major themes in the previous chapters. You might want to disable the router/firewall’s DHCP server capabilities to experiment with your new Linux server. Pay special attention to your 255. Give your DHCP client a static IP address from the same range that the DHCP server is supposed to provide. See whether you can ping the DHCP server.254.qxd 2/25/05 9:59 AM Page 140 140 Configuring the DHCP Server Chapter 8 DHCP Clients Obtaining 169. .0 Addresses Whenever Microsoft DHCP clients are unable to contact their DHCP server they default to selecting their own IP address from the 169.0.09Harrison_ch08. double-check your cabling and your NICs. Here are some steps you can go through to resolve the problem: 1.255 route.254.0 network until the DHCP server becomes available again. This is frequently referred to as Automatic Private IP Addressing (APIPA). CONCLUSION In most home-based networks. especially if your DHCP server has multiple interfaces. A DHCP server may be invaluable in an office environment where the time and cost of getting a network engineer to get the work done may make it simpler for Linux systems administrators to do it by themselves. a DHCP server isn’t necessary because the DSL router/firewall usually has DHCP capabilities.0.

ADDING USERS One of the most important activities in administering a Linux box is the addition of users. Who Is the Super User? The super user with unrestricted access to all system resources and files in Linux is the user named root. which is universally identified by Linux applications as belonging to a user with supreme privileges. It is not intended to be comprehensive. 141 . You need to log in as user root to add new users to your Linux server.10Harrison_ch09.qxd 2/25/05 10:01 AM Page 141 C H A P T E R 9 Linux Users and sudo IN THIS CHAPTER ☞ Adding Users ☞ Using sudo ☞ Conclusion Before we proceed. Here you’ll find some simple examples to provide a foundation for future chapters. but is a good memory refresher. You can use the command man useradd to get the help pages on adding users with the useradd command or the man usermod to become more familiar with modifying users with the usermod command. This user has a user ID of 0. it would be best to cover some basic user administration topics that will be very useful in later chapters.

You are prompted once for your old password and twice for the new one: ☞ User root changing the password for user paul: . The directory name will be the same as the user’s username: [root@bigboy tmp]# ll /home drwxr-xr-x 2 root root lost+found drwx-----2 accounts soho drwx-----2 alice children drwx-----2 derek children drwx-----2 jane parents drwx-----2 paul parents drwx-----2 sales soho sales[root@bigboy tmp]# 12288 Jul 24 20:04 1024 1024 1024 1024 1024 1024 Jul Jul Jul Jul Jul Jul 24 24 24 24 24 24 20:33 20:33 20:33 20:33 20:33 20:33 accounts alice derek jane paul How to Change Passwords You need to create passwords for each account. Add the Linux users and assign them to their respective groups: [root@bigboy [root@bigboy [root@bigboy [root@bigboy [root@bigboy [root@bigboy tmp]# tmp]# tmp]# tmp]# tmp]# tmp]# useradd useradd useradd useradd useradd useradd -g -g -g -g -g -g parents paul parents jane children derek children alice soho accounts soho sales If you don’t specify the group with the -g. Each user’s personal directory is placed in the /home directory. children. 4.10Harrison_ch09. read through these steps before starting: 1. Arrange your list of users into groups by function. this is also known as the User Private Group Scheme. Add the Linux groups to your server: [root@bigboy tmp]# groupadd parents [root@bigboy tmp]# groupadd children [root@bigboy tmp]# groupadd soho 3. they are prompted for their new permanent password. When each new user first logs in. Red Hat/Fedora Linux creates a group with the same name as the user you just created. and soho: Parents Paul Jane Children Alice Derek Soho Accounts Sales 2. This is done with the passwd command.qxd 2/25/05 10:01 AM Page 142 142 Linux Users and sudo Chapter 9 How to Add Users Adding users takes some planning. In this example there are three groups: parents.

maximum of 8 characters) Please use a combination of upper and lower case letters and numbers. New password: your new password Re-enter new password: your new password Password changed.10Harrison_ch09. Here is how unprivileged user paul would change his own password: [paul@bigboy paul]$ passwd Changing password for paul Old password: your current password Enter the new password (minimum of 5. [paul@bigboy paul]$ How to Delete Users The userdel command is used to remove the user’s record from the /etc/passwd and /etc/shadow used in the login process.qxd 2/25/05 10:01 AM Page 143 Adding Users 143 [root@bigboy root]# passwd paul Changing password for user paul. The data in a user’s directory can often be important even after the person has left your company: [root@bigboy tmp]# userdel -r paul How to Tell the Groups to Which a User Belongs Use the groups command with the username as the argument: [root@bigboy root]# groups paul paul : parents [root@bigboy root]# . [root@bigboy root]# ☞ Users might want to change their passwords at a future date. Use this option with care. New password: your new password Retype new password: your new password passwd: all authentication tokens updated successfully. the username: [root@bigboy tmp]# userdel paul There is also an optional -r switch that additionally removes all the contents of the user’s home directory. The command has a single argument.

USING SUDO If a server needs to be administered by a number of people. In the next example we change the ownership of the file named text.txt -rw-r--r-.1 root root 0 Nov 17 22:14 test.txt [root@bigboy tmp]# ll test.txt -rw-r--r-.txt [root@bigboy tmp]# chown testuser:users test.txt from being owned by user root and group root to being owned by user testuser in the group users: [root@bigboy tmp]# ll test.1 testuser users 0 Nov 17 22:14 test.txt [root@bigboy tmp]# You can also use the chown command with the -r switch for it to do recursive searches down into directories to change permissions. Example of a User Using sudo In this example. you will be prompted for your regular password before it is executed. which is an action that normally requires privileged access. All commands run as sudo are logged in the log file /var/log/messages. Without sudo. and where if everyone logs in with the same credentials.10Harrison_ch09. The sudo utility was designed to overcome this difficulty. the command fails: [bob@bigboy bob]$ more /etc/sudoers /etc/sudoers: Permission denied [bob@bigboy bob]$ . user bob attempts to view the contents of the /etc/sudoers file. when. The commands can be run as user root or as any other user defined in the /etc/sudoers configuration file. This is because it becomes difficult to determine exactly who did what.qxd 2/25/05 10:01 AM Page 144 144 Linux Users and sudo Chapter 9 How to Change the Ownership of a File You can change the ownership of a file with the chown command. When running the command with the sudo prefix. The sudo utility allows users defined in the /etc/sudoers configuration file to have temporary access to run commands they would not normally be able to due to file permission restrictions. The privileged command you want to run must first begin with the word sudo followed by the command’s regular syntax. it is normally not a good idea for them all to use the root account. You may run other privileged commands using sudo within a fiveminute period without being re-prompted for a password. The first argument is the desired username and group ownership for the file separated by a colon (:) followed by the filename.

1 Format of the /etc/sudoers File General sudoers File Record Format usernames/group servername = (usernames command can be run as) command There are some general guidelines when editing this file: ☞ Groups are the same as user groups and are differentiated from regular users by a % at the beginning.. ☞ You can have multiple usernames per line. . The visudo Command The visudo command is a text editor that mimics the vi editor that is used to edit the /etc/sudoers configuration file. visudo uses the same commands as the vi text editor.. General /etc/sudoers Guidelines The /etc/sudoers file has the general format shown in Table 9. Downloading and Installing the sudo Package sudo are covered in later Fortunately the package is installed by default by Red Hat/Fedora. . Table 9. The visudo command must run as user root and should have no arguments: [root@bigboy tmp]# visudo The /etc/sudoers File The /etc/sudoers file contains all the configuration and permission parameters needed for sudo to work. It is not recommended that you use any other editor to modify your sudo parameters because the sudoers file isn’t located in the same directory on all versions of Linux... which eliminates the need to anything more in this regard.10Harrison_ch09. There are a number of guidelines that need to be followed when editing it with visudo.qxd 2/25/05 10:01 AM Page 145 Using sudo 145 Bob tries again using sudo and his regular user password and is successful: [bob@bigboy bob]$ sudo more /etc/sudoers Password: . [bob@bigboy bob]$ The details of configuring and installing sections.1. separated by commas. The Linux user group users would be represented by %users.

10Harrison_ch09. ☞ If you run out of space on a line. ☞ The NOPASSWD keyword provides access without prompting you for your password. Simple /etc/sudoers Examples This section presents simple examples of how to do many commonly required tasks using the sudo utility. In most cases. /usr/local/apps/check. commands. /usr/sbin. and therefore offers the option to specify the names of servers. Granting Access to Specific Users to Specific Files This entry allows user peter and all the members of the group operator to gain access to all the program files in the /sbin and /usr/sbin directories. ☞ sudo assumes that the sudoers file will be used networkwide. . The example on using aliases in the sudoers file shows how to eliminate this problem. groups. Granting All Access to Specific Users You can grant users this sudoers entry: bob. ☞ The keyword ALL can mean all usernames. This is explained further in the next example. Notice how the trailing slash (/) is required to specify a directory location: peter. and servers. plus the privilege of running the command /usr/local/apps/check. bunny bob and bunny full access to all privileged commands with ALL=(ALL) ALL This is generally not a good idea because this allows bob and bunny to use the su command to grant themselves permanent root privileges. %operator ALL= /sbin/. Spaces are considered part of the command.qxd 2/25/05 10:01 AM Page 146 146 Linux Users and sudo Chapter 9 ☞ Multiple commands also can be separated by commas.pl. the file is used by only one server and the keyword ALL suffices for the servername. thereby bypassing the command logging features of sudo. you can end it with a back slash (\) and continue on the next line.pl Notice also that the lack of any username entries within parentheses () after the = sign prevents the users from running the commands automatically masquerading as another user.1. which will be using it in the servername position in Table 9.

bob. \ /usr/bin/ksh. /usr/bin/kill /usr/bin/pkill User peter is allowed to stop the monthend process with this command: [peter@bigboy peter]# sudo -u accounts pkill monthend Granting Access Without Needing Passwords This example allows all users in the group operator to execute all the commands in the /sbin directory without the need for entering a password. The sudoers file allows users to be grouped according to function with the group and then be assigned a nickname or alias. The sudoers entry would look like this: peter ALL=(accounts) /bin/kill. but first you have to be granted this privilege in the sudoers file. In the next example. users peter. bunny. \ /usr/bin/rsh. From time to time the application fails. This has the added advantage of being more convenient to the user: %operator ALL= NOPASSWD: /sbin/ Using Aliases in the sudoers File Sometimes you need to assign random groupings of users from various departments very similar sets of privileges. programmer peter is on the team developing a financial package that runs a program called monthend as user accounts. bob. Users ADMINS are then denied the option of running any SHELLS commands and su: Cmnd_Alias SHELLS = /usr/bin/sh.10Harrison_ch09. /usr/local/bin/zsh User_Alias ADMINS ADMINS = peter.qxd 2/25/05 10:01 AM Page 147 Using sudo 147 Granting Access to Specific Files as Another User The sudo -u entry enables you to execute a command as if you were another user. For example. !SHELLS . which is used throughout the rest of the file. Groupings of commands can also be assigned aliases. /usr/local/bin/tcsh. or /usr/bin/pkill commands but only as user accounts. %operator ALL = !/usr/bin/su. This feature can be convenient for programmers who sometimes need to kill processes related to projects they are working on. /usr/bin/kill. and bunny and all the users in the operator group are made part of the user alias ADMINS. All the command shell programs are then assigned to the command alias SHELLS. requiring peter to stop it with the /bin/kill. /usr/bin/csh.

or enter command shells that bypass sudo’s command logging.10Harrison_ch09. [root@bigboy tmp]# grep sudo /var/log/messages Nov 18 22:50:30 bigboy sudo(pam_unix)[26812]: authentication failure. the MySQL database application runs as user mysql and the Apache Web server application runs as user apache. but the restrictions can be enforced only as part of the company’s overall security policy. . Finally. All the sudo log entries have the word sudo in them. The advantage of this is that it helps to create an audit trail. You can even give some groups of users only partial access to privileged commands depending on their roles in the organization. which can be very helpful in determining how user error might have contributed to a problem. It doesn’t prevent them from copying the files to other locations to be run. COMMAND=/bin/more sudoers [root@bigboy tmp]# CONCLUSION It is important to know how to add users. PWD=/etc . so you can easily get a thread of commands used by using the grep command to filter the output accordingly. not just so they can log in to our system. Here is sample output from a user bob failing to enter his correct sudo password when issuing a command. Other Examples You can view a comprehensive list of command man sudoers. This makes sudo a valuable part of any company’s server administration and security policy. logname=bob uid=0 euid=0 tty=pts/0 ruser= rhost= user=bob Nov 18 22:51:25 bigboy sudo: bob : TTY=pts/0 . especially if the software is installed using TAR files. These accounts aren’t always created automatically.qxd 2/25/05 10:01 AM Page 148 148 Linux Users and sudo Chapter 9 This attempts to ensure that users don’t permanently su to become root. USER=root . Using syslog to Track All sudo Commands /etc/sudoers file options by issuing the All sudo commands are logged in the log file /var/log/messages. Most server-based applications usually run via a dedicated unprivileged user account—for example. immediately followed by the successful execution of the command /bin/more sudoers. the sudo utility provides a means of dispersing the responsibility of systems management to multiple users.

but this simple explanation should be enough: 149 . This chapter describes how you can make your Linux box into a Windows Primary Domain Controller (PDC) or a server for a Windows Workgroup.qxd 2/25/05 10:02 AM Page 149 C H A P T E R 10 Windows. What’s the difference between a PDC and Windows Workgroup member? A detailed description is beyond the scope of this chapter. such as printers. Linux. with Windows boxes. and Samba IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install Packages How to Get Samba Started The Samba Configuration File How SWAT Makes Samba Simpler Creating a Starter Configuration Samba Passwords How to Create a Samba PDC Administrator User How to Add Workstations to Your Samba Domain How to Add Users to Your Samba Domain Domain Groups and Samba How to Delete Users from Your Samba Domain How to Modify Samba Passwords Conclusion Samba is a suite of utilities that allows your Linux box to share files and other resources.11Harrison_ch10. Either configuration will allow everyone at home to have: ☞ Their own logins on all the home Windows boxes while having their files on the Linux box appear to be located on a new Windows drive ☞ Shared access to printers on the Linux box ☞ Shared files accessible only to members of their Linux user group.

Samba domains and Linux share the same usernames so you can log into the Samba-based Windows domain using your Linux password and immediately gain access to files in your Linux user’s home directory. You can create your own smb. For added security. which is easier. “Installing RPM Software. Keep in mind. Linux functionality doesn’t disappear when you do this. and with every client request. and Samba Chapter 10 ☞ A PDC stores the login information in a central database on its hard drive.conf with a text editor then subsequently use SWAT to edit the file. copy the file back to its original location. each PC stores the usernames and passwords locally so that they are unique for each PC. the Samba daemon reads the configuration file /etc/samba/smb. Tip As your smb. This allows each user to have a universal username and password when logging in from all PCs on the network. I’ll explain how to use both SWAT and a text editor to configure Samba later in this chapter.11Harrison_ch10. Samba mimics a Windows PDC in almost every way needed for simple file sharing. The files are named ☞ ☞ ☞ ☞ samba samba-common samba-client samba-swat When searching for the file. that if you create /etc/samba/smb. you’re better off editing a copy of it if you decide not to use SWAT. ☞ In a Windows Workgroup. you can make your Samba and Linux passwords different. however. test the validity of the changes using the testparm utility outlined in Chapter 12. Linux. . By default. Chapter 6. remember that the RPM’s filename usually starts with the RPM name followed by a version number. When it starts up.qxd 2/25/05 10:02 AM Page 150 150 Windows. If you need a refresher. “Samba Security and Troubleshooting.conf to determine its various modes of operation. Downloading and installing RPMs isn’t hard.0.” covers how to do this in detail. DOWNLOAD AND INSTALL PACKAGES Most Red Hat and Fedora Linux software products are available in the RPM format. After completing your modifications.i386. This chapter will cover only the much more popular PDC methodology used at home. you will lose all the comments you inserted with the text editor.0-15.conf is constantly being accessed. as in samba-client3.conf using a text editor or the Web-based SWAT utility. Samba is comprised of a suite of RPMs that come on the Fedora CDs.” and when you are satisfied with your changes.

) A share for storing domain logon information such as favorites and desktop icons.conf Description General Samba configuration parameters. A share for storing logon scripts.qxd 2/25/05 10:02 AM Page 151 The Samba Configuration File 151 HOW TO GET SAMBA STARTED You can configure Samba to start at boot time using the chkconfig command: [root@bigboy tmp]# chkconfig smb on You can start. Used for configuring printers. It is split into five major sections. Defines treatment of user logins. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep smb THE SAMBA CONFIGURATION FILE The /etc/samba/smb. Table 10. stop.1 File Format of smb.1 outlines. (Not created by default. . Samba does not need to be restarted after changes have been made to its configuration file. (Not created by default.11Harrison_ch10. and restart Samba after boot time using the tialization script as in the examples below: [root@bigboy tmp]# service smb start [root@bigboy tmp]# service smb stop [root@bigboy tmp]# service smb restart smb ini- Note Unlike many Linux packages.conf file is the main configuration file you’ll need to edit. You can test whether the smb process is running with the pgrep command.) Section [global] [printers] [homes] [netlogon] [profile] You can edit this file by hand or more simply through Samba’s SWAT Web interface. as it is read after the receipt of every client request. which Table 10.

each parameter box has its own online help. Each SWAT screen is actually a form that covers a separate section of the smb. and Samba Chapter 10 HOW SWAT MAKES SAMBA SIMPLER SWAT. you could save the original file with the name /etc/samba/smb.conf. Figure 10.qxd 2/25/05 10:02 AM Page 152 152 Windows. For ease of use.conf /etc/samba/smb.1 Samba SWAT main menu.conf file without needing to remember all the formatting.1 shows the main SWAT login screen. Linux.conf file into which you fill in the desired parameters. Basic SWAT Setup You must always remember that SWAT edits the smb. Samba’s Web-based configuration tool.original. as in [root@bigboy tmp]# cp /etc/samba/smb.conf file but also strips out any comments you may have manually entered into it beforehand. enables you to configure your smb.11Harrison_ch10. you should save a copy as a reference before proceeding with SWAT.original . For example.conf file has many worthwhile comments in it. Figure 10.conf. The original Samba smb.

This can automatically be switched between yes and no.0. because it eliminates these comments. ☞ The default configuration allows SWAT Web access from the VGA console only as user root on port 901 with the Linux root password.168.100 http://192.3 and localhost.168. using SWAT requires some understanding of the smb. you might want to create a Samba administrator user that has no root privileges or only enable SWAT access from the GUI console or localhost. Here is a sample: service swat { port socket_type protocol wait user server log_on_failure disable only_from } = 901 = stream = tcp = no = root = /usr/sbin/swat += USERID = no = localhost The file’s formatting is fairly easy to understand.3 bigboy IP with address 192.168. You may have to adjust your rules for this traffic to pass. Because this could be a security concern in a corporate environment. . “Telnet.1.1.conf file parameters. Notice that there are no commas between the entries. starting and stopping of SWAT is controlled by xinetd.d/swat. and xinetd. as you will see later. which is covered in Chapter 16. Therefore in this case you can also configure Samba on your Linux server from PC 192.168.0:901 in your browser to get the login screen. especially as there are only two entries of interest. ☞ The disable parameter must be set to no to accept connections.” via a configuration file named /etc/xinetd. only_from = localhost 192.1.qxd 2/25/05 10:02 AM Page 153 How SWAT Makes Samba Simpler 153 As you can see. This means you’ll have to enter http://127. TFTP. You can make SWAT accessible from other servers by adding IP address entries to the only_from parameter of the SWAT configuration file.0.168.11Harrison_ch10. SWAT doesn’t encrypt your login password.100:901. Become familiar with the most important options in this file before proceeding with SWAT. Here’s an example of an entry to allow connections only from 192.1. The enabling and disabling.1.3 using the URL Remember that most firewalls don’t allow TCP port 901 through their filters.

your name or your server’s hostname) []: 172. To activate SWAT use [root@bigboy tmp] chkconfig swat on To deactivate SWAT use [root@bigboy tmp] chkconfig swat off Encrypting SWAT By default SWAT is configured via an unencrypted Web link using the Linux root account.” for more details on encryption methods. [root@bigboy tmp]# cd /usr/share/ssl/certs [root@bigboy certs]# make stunnel. Remember to repeat this process next year. the chkconfig command automatically modifies the disable field accordingly in the configuration file and activates the change. you can create the encryption key certificate using the make command. and Samba Chapter 10 Controlling SWAT As with all xinetd-controlled applications.200 .1. .11Harrison_ch10.. but make sure you use the server’s IP address when prompted for your server’s Common Name or hostname. Common Name (eg..16.qxd 2/25/05 10:02 AM Page 154 154 Windows. (Please refer to the VPN section of Appendix I. Use all the defaults when prompted. [root@bigboy certs]# The resulting certificate has only a 365 day lifetime. You can configure SWAT to work only with securely encrypted HTTP (HTTPS) versus the regular HTTP method shown above. When running SWAT in the unsecured mode above you should take the added precaution of using it from the Linux console whenever possible.. “Miscellaneous Linux Topics. Linux..) Create an stunnel User You can create a stunnel user via the useradd command: [root@smallfry tmp]# useradd stunnel Create the Certificates From the /usr/share/ssl/certs directory. Here is how it’s done.pem .

root and the stunnel user.conf file to intercept SSL traffic on the SWAT port 901 and funnel it decrypted to a SWAT daemon running on port 902. Here’s how: # Configure stunnel to run as user “stunnel” placing temporary # files in the /home/stunnel/ directory chroot = /home/stunnel/ pid = /stunnel. Use the [root@bigboy certs]# chmod 640 stunnel.pem # Accept SSL connections on port 901 and funnel it to # port 902 for swat. you can configure the /etc/stunnel/stunnel. [swat] accept = 901 connect = 902 .conf Configuration File You can configure the stunnel application to: ☞ Intercept encrypted SSL traffic received on any TCP port ☞ Decrypt this traffic ☞ Funnel the unencrypted data to any application listening on another port.qxd 2/25/05 10:02 AM Page 155 How SWAT Makes Samba Simpler 155 Modify Certificate File Permissions The certificate needs to only be read by chmod and chgrp commands to do this.pid setuid = stunnel setgid = stunnel # Log all stunnel messages to /var/log/messages debug = 7 output = /var/log/messages # Define where the SSL certificates can be found. Client = no cert = /usr/share/ssl/certs/stunnel.pem [root@bigboy certs]# Create an /etc/stunnel/stunnel.pem [root@bigboy certs]# chgrp stunnel stunnel.1 root stunnel 1991 Jul 31 21:50 stunnel.11Harrison_ch10. For example.pem key = /usr/share/ssl/certs/stunnel.pem [root@bigboy certs]# ll /usr/share/ssl/certs -rw-r----.

it ignores the configuration file.d]# chkconfig swat off [root@bigboy xinetd.d directory.0. and Samba Chapter 10 Create a New /etc/xinetd.d]# chkconfig swat-stunnel on Start stunnel chkconfig com- Now start stunnel for the encryption to take place: [root@bigboy xinetd. If the daemon doesn’t find a match. Also make sure that the service is set to swat-stunnel.1 = root = /usr/sbin/swat += USERID = no = 127.0. copy the swat file and name it swat-stunnel. listening on port 902 and accepting connections only from localhost.0.0.1 Edit the /etc/services File to Create a Secure SWAT Entry The xinetd daemon searches the /etc/services file for ports and services that match those listed in each configuration file in the /etc/xinetd.11Harrison_ch10. [root@bigboy xinetd. You now have to edit /etc/services to include your new swat-stunnel file: swat-stunnel 902/tcp # Samba Web Administration Tool (Stunnel) Activate swat-stunnel You can then start the new swat-stunnel application with the mand. [root@bigboy certs]# cd /etc/xinetd.d]# cp swat swat-stunnel Your new swat-tunnel file should look like this: service swat-stunnel { port socket_type wait only_from user server log_on_failure disable bind } = 902 = stream = no = 127. Configure the new file to be enabled.d [root@bigboy xinetd.d]# stunnel . Linux.qxd 2/25/05 10:02 AM Page 156 156 Windows. You’ll also need to shutdown regular SWAT beforehand.d File for Secure SWAT To start.

.d]# netstat -an .qxd 2/25/05 10:02 AM Page 157 How SWAT Makes Samba Simpler 157 In Fedora Core 2.11Harrison_ch10.0..0:901 tcp 0 0 127. restart xinetd. stunnel should begin to function correctly. tcp 0 0 0.0.d]# rpm -e hwcrypto You then have to stop stunnel.0:* 0..d]# pkill stunnel [root@bigboy xinetd.. You need to uninstall hwcrypto to get stunnel to work correctly: [root@bigboy xinetd.not software-based encryption. and start stunnel again.0:* LISTEN LISTEN Test the Secure SWAT Login Point your browser to the Samba server to make an HTTPS connection on port 901: https://server-ip-address:901/ You will be prompted for the Linux root user username and password. you may get a cryptonet error when starting stunnel.0.d]# 0. .0.. Expect a delay of about 60 to 75 seconds with each login. After this.0. [root@bigboy xinetd. . Unfortunately stunnel doesn’t have a startup script in the /etc/init. The server will accept remote connections on port 901 only. [root@bigboy xinetd.d]# stunnel Test Secure SWAT Your Samba server should now be listening on both port 901 and 902 as shown in the netstat -an command that follows..0.0.0.d directory and needs to be terminated manually using the pkill command..:902 . as in: Unable to open “/dev/cryptonet” This is caused by an incompatibility with the hwcrypto RPM used for hardware. [root@bigboy xinetd. ..

there is no stunnel script in the /etc/init. servers.d directory to easily stop and restart it. so I’ll walk you through what you’ll find in each. Makes the PDC act as the central store for the names of all Windows clients.conf Value Yes Yes Description Tells Samba to become the PDC.2 explains the parameters you need to set to create a PDC.qxd 2/25/05 10:02 AM Page 158 158 Windows. Tells Samba to become the master browser across multiple networks all over the domain. Very helpful when you need to browse your local network for resources. and Samba Chapter 10 Troubleshooting Secure SWAT Sometimes you’ll make mistakes in the stunnel.conf file. CREATING A STARTER CONFIGURATION The easiest way to configure a Samba server to be the PDC for a small network is by using SWAT. Also known as a local master browser. You can also refer to Chapter 4. “Simple Network Troubleshooting. The [global] Section The [global] section governs the general Samba settings. The local master browsers register themselves with the domain master to learn about resources on other networks. stunnel doesn't start automatically when the system reboots because it doesn't have its own startup script.local file for this to occur. and printers on the network. You’ll need to edit the various sections of the smb. How to Make SWAT Changes Immediate SWAT immediately changes the functioning of Samba whenever you commit your changes through the Web GUI. Unfortunately. Parameter domain logons preferred master domain master Yes .2 Minimum Settings for the [global] Section of smb. You have to use the pkill command to stop it and the stunnel command to start it again: [root@bigboy tmp]# pkill stunnel . Linux. Table 10.11Harrison_ch10.” to isolate connectivity issues between the SWAT client and Samba server on TCP port 901 amongst other things. Table 10. but changes to this file take effect only after stunnel has been restarted. stunnel Make sure the file permissions and ownership on the stunnel. You will have to add the command stunnel in the /etc/rc. As mentioned previously.pem file are correct and that SWAT is always permanently off but swat-stunnel is permanently on.conf file.

User Makes domain logins query the Samba password database located on the Samba server itself. Here’s how to set the values using SWAT: 1. it keeps track of the IP addresses of all the domain’s servers and clients. You can view the contents of the configuration file by logging into the Samba server via a command prompt and using cat /etc/samba/smb. and some of the [global] parameters mentioned previously will be set to Auto.11Harrison_ch10. I’ve decided to use homenet. even though you may see them in SWAT.conf file. [global] workgroup = HOMENET time server = Yes domain logons = Yes os level = 65 preferred master = Yes domain masterc = Yes Be aware: Security [eq] user and WINS support [eq] yes are default settings for Samba. Log into SWAT. 4.conf to verify your changes as you do them. However the defaults may not be to your liking. A value of 65 usually makes the Samba server win.qxd 2/25/05 10:02 AM Page 159 Creating a Starter Configuration 159 Parameter Os level Value 65 Description wins support time server domain security Sets the priority the Samba server should use when negotiating to become the PDC with other Windows servers. and click on the [global] section. and they may not show up in your smb. Click the Advanced button to see all the options. Your smb. the default domain is MYGROUP. Yes Allows the Samba server to provide name services for the network. 3. Using the SWAT Wizard The SWAT utility has a Wizard button that can be used to configure your server as a PDC quickly. . The name you select is your choice. “homenet” The name of the Windows domain you’ll create.conf file should resemble the example below when you’re finished. In other words. and click on the Commit Changes button when finished. 2. Make your changes. For example. Yes Lets the Samba server provide time updates for the domain’s clients.

Table 10. added to your PC. Table 10. Makes new files created by the user to have 644 permissions. This is actually a virtual drive that maps to the corresponding Linux users’ login directories on the Linux PDC. you’ll see a new drive. Parameter browseable read only create mask directory mask 0775 Here’s how to set the values using SWAT: 1.conf to verify your changes as you do them. You want to change this to 0700 so that only the login user has access to directories. Click on the SWAT shares button to proceed to where shared directories are configured. Make your changes. Samba considers all directories to be shares that can be configured with varying degrees of security. Allows Samba users to also write to their Samba Linux directories.conf Value No No 0664 Description Doesn’t allow others to browse the contents of the directory. 3. Makes new subdirectories created by the user to have 775 permissions. Your smb. Click the Advanced button to see all the options. [homes] read only = No browseable = No create mask = 0644 directory mask = 0755 .qxd 2/25/05 10:02 AM Page 160 160 Windows. 2. Choose the [homes] share section.3 Minimum Settings for [home] Section of smb. and Samba Chapter 10 The [homes] Section Part of the process of adding a user to a Samba domain requires you to create a Linux user on the Samba PDC itself. and click on the Commit Changes button when finished. You can view the contents of the configuration file by logging into the Samba server via a command prompt and using cat /etc/samba/smb. When you log into the Samba PDC.conf file should resemble this when finished.11Harrison_ch10.3 explains the minimum settings you need to create a functional [homes] section. usually named Z:. The [homes] section governs how Samba handles default login directories. 4. You want to change this to 0600 so that only the login user has access to files. Linux.

There is also a share under [printers] called printers that governs common printer settings. Create a [netlogon] share section. 4. 5. Click on the Commit Changes button. Click on the Commit Changes button. 6. Remember to create these share directories from the command line afterwards: [root@bigboy tmp]# mkdir -p /home/samba/netlogon [root@bigboy tmp]# mkdir -p /home/samba/profile [root@bigboy tmp]# chmod -r 0755 /home/samba The [printers] Share Section Samba has special shares just for printers. and read only settings. Create a [profiles] share section. The [profiles] share section stores settings related to the look and feel of Windows so that the user has the same settings no matter which Windows PC is logged into. The mask settings allow only the owner of the netlogon subdirectory to be able to modify its contents. mask. Click the Shares button. The default smb. Print shares always have the printable parameter set to yes. 3. Modify the path and guest ok settings. and these are configured in the [printers] section of SWAT.11Harrison_ch10.qxd 2/25/05 10:02 AM Page 161 Creating a Starter Configuration 161 The [netlogon] and [profiles] Share Sections The [netlogon] share section contains scripts that the Windows clients may use when they log into the domain. Modify the path. 7.conf file should look like this when you’re finished: [netlogon] path = /home/samba/netlogon guest ok = Yes [profiles] path = /home/samba/profiles read only = No create mask = 0600 directory mask = 0700 Here’s how to do it: 1. 2. Your smb.conf [printers] share section looks like this: . The [profiles] share section stores such things as favorites and desktop icons.

HOW TO CREATE A SAMBA PDC ADMINISTRATOR USER To do both SWAT and user administration with Samba. Home Environment By default. the root user is the Samba administrator.qxd 2/25/05 10:02 AM Page 162 162 Windows. Use the passwd command to change Linux passwords. Fortunately. Samba domain logins use the smbpasswd password.conf file you saved at the beginning of this exercise has many varied examples that you may use and apply to your particular environment. Samba passwords are stored in the /etc/samba/smbpasswd file. you’ll need to create administrator accounts on the Samba PDC Linux server. and SWAT requires you to use the Linux root password.” SAMBA PASSWORDS You should be aware that your Linux password and Samba passwords are stored in two different locations. This provides the Samba administer the flexibility of allowing only some of the Linux users to have Samba accounts. Linux. you can add workstations to the Windows domain by creating a Samba-specific root password using the smbpasswd command. “Sharing Resources Using Samba. You can find the steps for creating a simple shared directory for home users in Chapter 11. and Samba Chapter 10 [printers] comment = All Printers path = /var/spool/samba printable = Yes browseable = No Shares for Specific Groups of Users The default Samba Version 3 smb.11Harrison_ch10. which are stored in the /etc/shadow file. and you can change them with the smbpasswd command. This difference is important. TELNET. as you will see throughout the chapter. [root@bigboy tmp]# /usr/bin/smbpasswd -a root password Note Remember that regular Linux logins via the console. or SSH require the Linux passwd command. Samba passwords are stored in the /etc/samba/smbpasswd file. .

qxd 2/25/05 10:02 AM Page 163 How to Add Workstations to Your Samba Domain 163 Corporate Environment In a corporate environment. Use the smbpasswd command to create Samba passwords for domain logins for this group. such as sysadmin. the commands disable the password for this user and add the machine added to the smbpasswd file to help keep track of which devices are members of the domain. 4. giving each an individual username. Create Samba Trust Accounts for Each Workstation PDCs accept user logins only from trusted PCs that have been placed in its PC client database. Create a Linux user group. (Remember that Linux passwords are changed with the passwd command.conf file so that the sysadmin group is listed in the [global] parameter settings: domain admin group = @sysadmin admin users = @sysadmin printer admin = @sysadmin 3. via telnet or ssh. and replace machine_name with the name of the Windows client machine. 2. In summary. Pay careful attention to the $ at the end. you might want more than one person to administer Samba. Use SWAT to update your smb. a machine trust account needs to have corresponding entries in the /etc/passwd and /etc/smbpasswd files. Manual Creation of Machine Trust Accounts (NT Only) The commands in the example create a special Linux group for Samba clients and then add a special machine user that’s a member of the group.) HOW TO ADD WORKSTATIONS TO YOUR SAMBA DOMAIN Adding workstations to a Samba domain is a two-step process involving the creation of workstation trust accounts on the Samba server and then logging into each workstation to add them to the domain. For security reasons this password may be different from the Linux password used to log into the Linux system from the console. with the groupadd command.11Harrison_ch10. Here are the steps to follow: 1. Samba can create these Machine Trusts either manually or automatically. [root@bigboy tmp]# groupadd samba-clients [root@bigboy tmp]# /usr/sbin/useradd -g samba-clients \ . Next. Create individual Linux users that are part of this group.

[root@bigboy tmp]# smbpasswd -a -m machine_name Samba version 3 adds it automatically. ... You can set this up by editing /etc/samba/smb. [global] # <. and Samba Chapter 10 -d /dev/null -s /bin/false machine_name$ [root@bigboy tmp]# passwd -l machine_name$ [root@bigboy tmp]# smbpasswd -a -m machine_name This is the only way to configure machine trusts using Windows NT. therefore.qxd 2/25/05 10:02 AM Page 164 164 Windows. The next sections show you how to add the most popular versions of Windows clients to your domain. [root@bigboy tmp]# groupadd samba-clients In Samba version 2. The easiest way is to use SWAT in the Global menu to modify the add machine script parameter. don’t require machine trust accounts. Dynamic Creation of Machine Trust Accounts Although you can use the manual method. you’ll need to create the samba-clients Linux group that will help identify all the domain’s Windows clients listed in the /etc/passwd file.. Linux. you need to add the client to the smbpasswd file also. which is known as making a machine account on the fly. the recommended way of creating machine trust accounts is simply to allow the Samba server to create them as needed when the Windows clients join the domain.11Harrison_ch10. and each version has its own procedure for joining a domain. Navigate to the Network section of the Control Panel (Start>Settings>Control Panel>Network). Windows 95/98/ME and Windows XP Home Windows 9x machines do not implement full domain membership and. Select the Configuration tab.. 2. Make Your PC Clients Aware of Your Samba PDC There are many types of Windows installed on PCs.> add machine script = /usr/sbin/useradd -d /dev/null -g samba-clients s /bin/false -M %u When you’ve completed the modifications.remainder of parameters.conf to automatically add the required users. Here’s what you need to do: 1.

You will be prompted for a user account and password with rights to join a machine to the domain. Click the Properties button. Press the Windows and Break keys simultaneously to access the System Properties dialog. You can now log in using any account in the /etc/smbpasswd file with your domain as the domain name. In this case. Enter your domain name and computer name. and then click OK. 5. 4. Reboot. Click on the Network Identification or Computer Name tab on the top. 2. create a dynamic Samba machine trust account. . You can now log in using any account in the /etc/smbpasswd file with your domain as the domain name. simply reboot and you’re done! Windows NT For Windows NT. Navigate to the Network section of the Control Panel (Start>Settings>Control Panel>Network). You should get a “Welcome to <DOMAIN>” message as confirmation that you’ve been added. Windows 200x and Windows XP Professional For the 200x and XP Professional varieties of Windows. In this home environment scenario. then follow these steps: 1. 6. Reboot. 5.qxd 2/25/05 10:02 AM Page 165 How to Add Workstations to Your Samba Domain 165 3. 3. the existing machine trust account joins the machine to the domain. Check Log onto Windows NT Domain. Now. Click all the OK buttons. the user would be root with the corresponding smbpasswd password. 7. Click the Change button. you must first create a manual Samba machine trust account as explained earlier. 6. and enter the domain name. 6. 4.11Harrison_ch10. Enter the domain name and computer name. Click on the Member of Domain button. then go through these steps: 1. 2. 5. you should get a “Welcome to <DOMAIN>” message confirming that you’ve been added. Enter the information for your Samba administrator. Select the Identification tab. do not check the box Create a Computer Account in the Domain. 4. 3. Click OK. Click the Properties button. Now. Highlight Client for Microsoft Networks.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 166

166

Windows, Linux, and Samba

Chapter 10

With Samba version 2, however, you may also have to make a few changes to your system’s registry using the regedit command:
[HKEY_LOCAL_MACHINE\SYSTEM\CurrentControlSet\Services\Netlogon\ Parameters] “requiresignorseal”=dword:00000000 “signsecurechannel”=dword:00000000

After making these changes, reboot before continuing.

HOW TO ADD USERS TO YOUR SAMBA DOMAIN
Adding users to a domain has three broad phases. The first is adding a Linux user on the Samba server, the second is creating a Samba smbpasswd that maps to the new Linux user, and the third is to map a Windows drive letter to the user’s Linux home directory. Take a closer look. Adding the Users in Linux First, go through the process of adding users in Linux just as you would normally. Passwords won’t be necessary unless you want the users to log into the Samba server via TELNET or SSH. To create the user, use the command
[root@bigboy tmp]# useradd -g 100 peter

Giving them a Linux password is necessary only if the user needs to log into the Samba server directly. If the user does, use this method:
[root@bigboy tmp]# passwd peter Changing password for user peter. New password: Retype new password: passwd: all authentication tokens updated successfully. [root@bigboy tmp]#

Mapping the Linux Users to an smbpassword Next, you need to create Samba domain login passwords for the user:
[root@bigboy tmp]# /usr/bin/smbpasswd -a username password

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 167

How to Add Users to Your Samba Domain

167

The -a switch adds the user to the /etc/smbpasswd file. Use a generic password then have users change it immediately from their workstations in the usual way. Remember the smbpasswd sets the Windows domain login password for a user, which is different from the Linux login password to log into the Samba box. Mapping a Private Windows Drive Share By default, Samba automatically gives each user logged into the domain an H: drive that maps to the /home/username directory on the Linux box.
Mapping Using My Computer

If the auto-mapping doesn’t work, then try: 1. Let the user log into the domain. 2. Right-click on the My Computer icon on the desktop. 3. Click on Map Network Drive. 4. Select a drive letter. 5. Browse to the HOMENET domain, then the Samba server, then the user’s home directory. 6. Click on the check box Reconnect at Logon, to make the change permanent. If you’re more a typist than a clicker, try the next method.
Mapping from the Command Line

If you find the My Computer method too time consuming for dozens of users or if the PC doesn’t have the feature available, then you can use the commandline method and possibly make it into a script: 1. Create a master logon batch file for all users.
[root@bigboy tmp]# vi /home/samba/netlogon/login.bat

2. Add the following lines to mount the user’s share as drive P: (for private):
REM Drive Mapping Script net use P: \\bigboy\

3. Make the file world readable using:
[root@bigboy tmp]# chmod 644 /home/samba/netlogon/login.bat

4. Linux and Windows format text files differ slightly. As the file resides on a Linux box, but will be interpreted by a Windows machine, you’ll have to convert the file to the Windows format. Use the unix2dos command:
[root@bigboy tmp]# unix2dos /home/samba/netlogon/login.bat unix2dos: converting file /home/samba/netlogon/login.bat to DOS format ... [root@bigboy tmp]#

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 168

168

Windows, Linux, and Samba

Chapter 10

5. The final step is to edit your smb.conf file’s [global] section to have a valid entry for the logon script parameter. This can be done using SWAT via the Globals menu.
[global] logon script = login.bat

Now your users will have additional disk space available on a Windows P: drive whenever they login.

DOMAIN GROUPS

AND

SAMBA

Samba supports domain groups that allow users who are members of the group to have administrator rights on each PC in the domain. This enables them to add software and configure network settings, among other tasks. In Windows, domain groups also have the ability to join machines to the domain; however, Samba does not support this currently. The domain admin group parameter specifies users who will have domain administrator rights. The argument is a space-separated list of user names or group names (group names must have an @ sign prefixed). For example:
domain admin group = USER1 USER2 @GROUP

HOW TO DELETE USERS

FROM YOUR

SAMBA DOMAIN

Deleting users from your Samba domain is a two-stage process in which you have to remove the user from the Linux server and also remove the user’s corresponding smbpasswd entry: 1. Delete the users using the smbpasswd with the -x switch:
[root@bigboy tmp]# smbpasswd -x john Deleted user john. [root@bigboy root]#

2. Delete the Linux user by following the normal deletion process. For example, to delete the user john and all john’s files from the Linux server, use:
[root@bigboy tmp]# userdel -r john

Sometimes you may not want to delete the user’s files so that they can be accessed by other users at some other time. In this case, you can just deactivate the user’s account using the passwd -l username command.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 169

Conclusion

169

HOW TO MODIFY SAMBA PASSWORDS
You can set your Samba server to allow users to make changes in their domain passwords and have these changes mirrored automatically in their Linux login passwords. Table 10.4 explains the [global] smb.conf parameters that you need to change.
Table 10.4 Parameter
unix passwd sync passwd program passwd chat

The smb.conf Settings for Enabling Online Password Changes Value Yes Use the SWAT defaults Use the SWAT defaults Description Enables Samba/Linux password synchronization. Lists the location of the Linux password file, which is usually /bin/passwd. A short script to change the Linux password using the Samba password.

CONCLUSION
By now you should have a fairly good understanding of adding users and PCs to a Samba domain. The next step is to take better advantage of Samba’s file sharing features, which is Chapter 11’s task.

11Harrison_ch10.qxd

2/25/05

10:02 AM

Page 170

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 171

C H A P T E R

11

Sharing Resources Using Samba

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ Adding a Printer to a Samba PDC Creating Group Shares in Samba Sharing Windows Drives Using a Linux Samba Client Automating Mounting with Linux Samba Clients Conclusion

Now that you have Samba up and running, you may want to allow users to share such resources as floppy drives, directories, and printers via the Samba server. This chapter tells you how to do it all.

ADDING

A

PRINTER TO A SAMBA PDC

Sharing printers amongst all your PCs is one of the advantages of creating a home network. Here’s how to connect your printer directly to your PDC. Not only does this method make your printer available to all your Windows workstations, it also makes your Samba PDC a print server! The only potential snag is that you need the Windows printer driver loaded on every client machine. This may be okay for a small home network but impractical for a huge corporate network. Adding the Printer to Linux By far, the easiest way to add a printer in Linux is to use one of the many menu-based printer utilities available. For the example, I’ll use systemconfig-printer (Figure 11.1), which is easy to find and can be accessed from the command line:
[root@bigboy tmp]# system-config-printer

171

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 172

172

Sharing Resources Using Samba

Chapter 11

Figure 11.1

Printer configuration screen.

Note Sometimes the graphics in system-config-printer don’t work as expected due to your environment variables. You can temporarily set the required variable to the correct output with the command
[root@bigboy tmp]# env LANG=C system-config-printer

Assuming your printer is locally attached to the parallel port, here are the steps to use: 1. Using the Tab key, move to the New button and press the Enter key. 2. When the Create a New Queue menu appears, give the printer a name that’s easy to remember, select Local Printer Device, and move to the “next” button before pressing Enter. 3. From the resulting Setting Up Local Printer Device menu, select /dev/lp0, assuming the printer is on the parallel port as opposed to the USB port. Click the Next button to go to the Queue Driver menu. 4. Scroll to your manufacturer’s entry, press Enter. Scroll further to your model, and press Enter again. You’ll now get a choice of drivers, select the

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 173

Adding a Printer to a Samba PDC

173

default device, which is marked with an asterisk. Click Next to go to the Create a New Queue confirmation menu. 5. Move to the Finish button, and press Enter. There will be a slight delay. At this stage, it’s wise to do a test print to make sure all is okay. Make Samba Aware of the Printer The easiest way to let Samba know the printer is available is via the Samba SWAT Web interface. Once you are in SWAT: 1. Click the Printers button. 2. Find your printer in the pull-down menu. If the printer name has an asterisk beside it, it has been auto-configured by Samba. It might not be visible on your network, however, if Samba hasn’t been restarted since creating the printer. If this is the case, restart Samba. You can now skip ahead to the “Configure the Printer Driver on the Workstations” section. 3. If Samba did not auto-configure your printer, you need to edit or create it yourself. Click on the Commit Changes button to create an updated /etc/samba/smb.conf file. 4. Click on the Status tab at the top of the screen, and restart smbd and nmbd to restart Samba. Your printer will now be available for use by all Windows workstations. Configure the Printer Driver on the Workstations With the printer ready to go on the Linux side, you now need to prepare things in Windows. 1. Download the appropriate Windows printer driver from the manufacturer and install it. 2. Go to the Add Printer menu. Click the Next button. 3. Select the Network Printer button to access the Local or Network Printer menu. Click the Next button, again. 4. You now should be on the Locate Your Printer menu. Don’t enter a name, instead click Next so you can browse for your printer. 5. From the Browse for Printer menu, double-click on the name of your Linux Samba server. You should see the new printer. Click on the printer name, then click Next. 6. You may get the message “The server on which the printer resides does not have the correct printer driver installed. If you want to install the

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 174

174

Sharing Resources Using Samba

Chapter 11

driver on your local computer, click OK.” Fortunately, you pre-installed the driver. Click the OK button. 7. When the Add Printer Wizard appears, select the manufacturer of your printer, select the printer model, and then click OK. 8. The Add Printer Wizard will ask you whether you want to use this new printer as the default printer. Select Yes or No depending on your preference. Click the Next button 9. From the resulting Completing the Add Printer Wizard menu, click the Finish button. The new printer should now show up on the Windows Printers menu in the Control Panel. Send a test print, to be sure all is well.

CREATING GROUP SHARES

IN

SAMBA

On occasion, subgroups of a family need a share that is fully accessible by all members of the group. For example, parents working in a home office environment may need a place where they can share, distribute, or collaboratively work on documents. Here’s how it’s done. Create the Directory and User Group As with any group activity, the first step is to get organized. 1. Create a new Linux group named parents:
[root@bigboy tmp]# /usr/sbin/groupadd parents

2. Create a new directory for the group’s files. If one user is designated as the leader, you might want to change the chown statement to make that person owner:
[root@bigboy tmp]# mkdir /home/parent-files [root@bigboy tmp]# chgrp parents /home/parent-files [root@bigboy tmp]# chmod 0770 /home/parent-files

3. Add the group members to the new group. For instance, the command to add a user named father to the group is
[root@bigboy tmp]# /usr/sbin/usermod -G parents father

All your members are in the group; now they need to share. Configure the Share in SWAT Next, you need to create the share in Samba using SWAT. 1. Click on the Shares button, then enter the name of the share you want to create, such as only-parents.

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 175

Sharing Windows Drives Using a Linux Samba Client

175

2. Click on the Create Share button. Make sure the path maps to /home/ parent-files, and make the valid users be @parents, where parents is the name of the Linux user group. 3. Click on the Commit Changes button to create a new /etc/samba/smb.conf file. 4. Click on the Status tab at the top of the screen and restart smbd and nmbd to restart Samba. Your /etc/samba/smb.conf file should have an entry like this at the end
# Parents Shared Area [only-parents] path = /home/parent-files valid users = @parents

If it does, all is well and you can move on. If not, double check your work in the last steps. Map the Directory Using My Computer Finally, let the user log into the domain from a remote PC: 1. 2. 3. 4. Right click on the My Computer icon on the desktop. Click on Map Network Drive. Select a drive letter. Browse to the HOMENET domain, then the Samba server, then the share named only-parents. 5. Click on the check box Reconnect at Logon, to make the change permanent. Now the files located in the Linux /home/parent-files directory will be accessible to the parents only and your job is complete!

SHARING WINDOWS DRIVES USING

A

LINUX SAMBA CLIENT

Up to this point I have focused on your Linux server being a Samba server, but it can also mimic a Windows client using Samba’s client software. For example, you can also access a CD-ROM, DVD, Zip, floppy, or hard drive installed on a Windows machine from your Linux box. In this section I’ll show you how to share a CD-ROM drive. Windows Setup The Windows client box should be set up first as a member of a Samba domain or workgroup. The next step is to make the CD-ROM drive shared. The steps you use depend on which version of Windows you have.

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 176

176

Sharing Resources Using Samba

Chapter 11

For Windows 98/ME: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing. Set the Share Name as cdrom with the appropriate access control. Restart Windows. For Windows 2000: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing. Set the Share Name as cdrom with the appropriate access control. Log out, and log in again as normal using your current login. Finally, for Windows XP: 1. 2. 3. 4. Double-click My Computer. Right-click on the CD-ROM drive, and choose Sharing and Properties. Set the Share Name as cdrom, and set the appropriate access control. Log out and log in again as normal using your current login.

After you have completed this task, you’ll have to go to the next step of testing your configuration. Test Your Windows Client Configuration Use the smbclient command to test your share. You should substitute the name of your Windows client PC for “WinClient,” and in place of “username” provide a valid workgroup/domain username that normally has access to the Windows client. You should get output like this when using the username’s corresponding password:
[root@bigboy tmp]# smbclient -L WinClient -U username added interface ip=192.168.1.100 bcast=192.168.1.255 nmask=255.255.255.0 added interface ip=127.0.0.1 bcast=127.255.255.255 nmask=255.0.0.0 Got a positive name query response from 192.168.1.253 ( 192.168.1.253 ) Password: Domain=[HOMENET] OS=[Windows 5.1] Server=[Windows 2000 LAN Manager] Sharename Type Comment --------- ---- ------IPC$ IPC Remote IPC D$ Disk Default share print$ Disk Printer Drivers SharedDocs Disk cdrom Disk Printer2 Printer Acrobat PDFWriter

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 177

Sharing Windows Drives Using a Linux Samba Client

177

ADMIN$ Disk Remote Admin C$ Disk Default share Server Comment --------- ------Workgroup Master --------- -------

You can get the same result with
[root@bigboy tmp]# smbclient -L WinClient -U username%password

but this method is less secure as your password is echoed on the screen. Create a CD-ROM Drive Mount Point on Your Samba Server You’ll now need to create the mount point on the Linux server to mount and access the CD-ROM drive. In this case, I’ve named it /mnt/winclient-cdrom, and you’ll use the mount command to get access to this device from the Linux server.
Password Prompt Method

The Linux mount command will try to access the CD-ROM device as user username by using the username[eq] option. You will be prompted for a password:
[root@bigboy tmp]# mkdir /mnt/winclient-cdrom [root@bigboy tmp]# mount -t smbfs -o username=username \ //winclient/cdrom /mnt/winclient-cdrom

No Prompt Method

Linux won’t prompt you for a password if you embed the access password into the mount command string along with username:
[root@bigboy tmp]# mkdir /mnt/winclient-cdrom [root@bigboy tmp]# mount -t smbfs -o \ username=username,password=password \ //winclient/cdrom /mnt/cdrom

The smbmount Command Method

Some versions of Linux support the smbmount command to mount the remote drive. Incompatible versions will give errors like this:
[root@bigboy tmp]# smbmount //winclient/cdrom \ /mnt/winclient-cdrom -o username=username

12Harrison_ch11.qxd

2/25/05

10:03 AM

Page 178

178

Sharing Resources Using Samba

Chapter 11

Password: 27875: session setup failed: ERRDOS - ERRnoaccess (Access denied.) SMB connection failed

To be safe, stick with using the Linux mount command.

AUTOMATING MOUNTING WITH LINUX SAMBA CLIENTS
You can also automate the mounting of shares by placing entries in your /etc/fstab file. In the example below the home directory of user peter on server 192.168.1.100 will be mounted on the /mnt/smb mount point as a Samba filesystem (smbfs) using the login information in the file named /etc/cred.
# # File: /etc/fstab # //192.168.1.100/peter

/mnt/smb

smbfs

credentials=/etc/cred 0 0

The contents of the /etc/cred file needs to have the username and password for the account in this format
# # File: /etc/cred # username = peter password = peterspassword

Once finished, you can use mount -a to mount the new /etc/fstab entry, and the /mnt/smb directory will contain the contents of the share.
[root@smallfry tmp]# mount -a [root@smallfry tmp]# ls /mnt/smb backups profile docs data music [root@smallfry tmp]#

CONCLUSION
Both this chapter and the last have detailed the steps needed to configure a Samba network that is adequate for a small office or home. There are many steps to take, none are particularly complex, but you run the risk of not getting Samba to work if you omit any of them. For this reason, the next chapter is a dedicated troubleshooting guide to help you diagnose and recover from the most common Samba mistakes that we all tend to make.

13Harrison_ch12. and one small slip up can have big repercussions.conf Processing section “[homes]” Processing section “[printers]” 179 . the initial configuration of Samba can be tricky. If you used SWAT to edit the file. Many simple steps need to be executed in the correct order.qxd 2/25/05 10:03 AM Page 179 C H A P T E R 12 Samba Security and Troubleshooting IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Testing the smb.conf File Samba and Firewall Software Testing Basic Client/Server Network Connectivity Testing Samba Client/Server Network Connectivity Checking the Samba Logs Samba Network Troubleshooting Basic Samba Security Conclusion Configuring Samba for your office or home can provide many advantages. This chapter explores the ways in which you can recover from those mistakes that you couldn’t avoid. in some cases. software installation and maintenance. Unfortunately.conf file. By encouraging users to store files on a central file server you can simplify data backup and. TESTING THE SMB. you will usually pass the test successfully. as shown: [root@bigboy tmp]# testparm -s Load smb config files from /etc/samba/smb.CONF FILE Samba has a test utility called testparm that alerts you to errors in the smb.

..138 -j DROP iptables -A INPUT -p tcp -d $SAMBA_SERVER -m multiport \ --dports 139.445 -j ACCEPT iptables -A INPUT -p udp -s $NETWORK -d $BROADCAST --dport 137 \ -j ACCEPT iptables -A INPUT -p udp -d $SAMBA_SERVER -m multiport \ --dports 137. You have two options for working with it.255. TCP ports 139) and TCP port 445 for SMB file sharing without NetBIOS. You can ensure that this is deactivated..conf configuration a file that’s different from the default file. which may be desirable on a secured network.100 “ NETWORK=”192.138 -j ACCEPT iptables -A INPUT -p tcp -s $NETWORK -d $SAMBA_SERVER -m multiport \ --dports 139.168. Here is sample script snippet: #!/bin/bash SAMBA_SERVER=”192. Or.1. [root@bigboy tmp]# A successful test only means that Samba will load the configuration file.168.. .445 -j DROP iptables -A OUTPUT -s $SAMBA_SERVER -d $NETWORK -m state --state \ ESTABLISHED. Simply provide the filename as the first SAMBA AND FIREWALL SOFTWARE Firewall software installed both on your Windows PCs and on the Samba server may prevent Samba from functioning.255” # Local area network Broadcast Address iptables -A INPUT -i lo -j ACCEPT iptables -A OUTPUT -o lo -j ACCEPT iptables -A INPUT -p udp -s $NETWORK -d $SAMBA_SERVER \ -m multiport --dports 137. Two popular packages—iptables and Zone Alarm—offer solutions. you can configure it to allow through such Microsoft protocols as NetBIOS (UDP ports 137 and 138..1. Linux iptables The Fedora installation process configures the iptables firewall package by default.RELATED -j ACCEPT .0/24” # Local area network BROADCAST=”192.168. Tip You can use argument: [root@bigboy tmp]# testparm -s filename testparm to test /etc/samba/smb.13Harrison_ch12. There are other causes for Samba problems.qxd 2/25/05 10:03 AM Page 180 180 Samba Security and Troubleshooting Chapter 12 Loaded services file OK.

168. The easiest way around this is to configure Zone Alarm to consider your home network as a trusted network too. and Zone Alarm should cease to interfere with Samba. To do so. and select Properties. 4. it will most likely have a 192. Click on the Advanced tab and then on the Windows Firewall Settings button. you may want to experiment with the firewall software settings to optimize your security. Double-click on the Network Connections icon. After you get SAMBA to work. therefore. This means that the software will deny all inbound connections that attempt to connect with your PC. 2. Follow these steps: 1. . instead of an Internet network.qxd 2/25/05 10:03 AM Page 181 Testing Basic Client/Server Network Connectivity 181 For more information. The NetBIOS traffic that Samba uses to communicate with the PCs on the network. Turn off the Internet Connection Firewall by clearing its check box. You may also leave the firewall on. but allow Windows file sharing traffic through this connection. please refer to Chapter 14. Make this network a trusted network. This can be done by clicking on the Exceptions tab of the Windows Firewall dialog box and clicking on the File and Printer Sharing check box.x/255. 5.13Harrison_ch12.255. 3. The Windows XP Built-In Firewall You may also need to disable the firewall feature of Windows XP. keeping in mind the need to maintain a valid relationship with the Samba server. click on the Firewall tab and edit the settings for your home network.” Windows-based Zone Alarm The default installation of Zone Alarm assumes that your PC is directly connected to the Internet. is considered as hostile traffic. Right-click on your LAN connection icon.0 type entry. Bring up the Control Panel. TESTING BASIC CLIENT/SERVER NETWORK CONNECTIVITY You can perform several tests to ensure that the Samba server and all its workstations can do basic communication with each other.255. “Linux Firewalls Using iptables.x.

0. Ping the client’s IP address.0. Next. Check your SWAT configuration for invalid hosts allow. Ping the client’s IP address and loopback address (127.0. TESTING SAMBA CLIENT/SERVER NETWORK CONNECTIVITY After configuring basic network connectivity. 3. and invalid users entries. Failure of this test may mean that Samba isn’t running on the server at all and may need to be started. your routing. Ping the server using its DNS name. check your cabling. hosts deny. Ping the server’s IP address and loopback address (127.1).2-7. 2. Ping the client using its DNS name. As part of a thorough troubleshooting procedure: 1. If any of these fail. 3. Ping the server’s IP address.qxd 2/25/05 10:03 AM Page 182 182 Samba Security and Troubleshooting Chapter 12 From the Samba server: 1.0. Press the Enter key when prompted for a password.0.FC1] Server --------SILENT Workgroup --------HOMENET OTHERNET [root@bigboy tmp]# Comment ------Samba Server Master ------BIGBOY SILENT .FC1] Sharename Type Comment -----------------IPC$ IPC IPC Service (Samba Server) ADMIN$ IPC IPC Service (Samba Server) Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3.13Harrison_ch12. or the presence of a firewall running on either the server or client. 2.0. Make sure your Samba server can see all the shares available on the network with the smbclient -l samba_server command.2-7. [root@bigboy tmp]# smbclient -L bigboy Password: Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3.1). you need to go through a variety of tests to determine whether Samba has been configured correctly both on the server and client. from the Samba client: 1.

103 *<00> [root@bigboy tmp]# 4. This test actually sends out a broadcasted request for information.0 added interface ip=192.1.103 192.168.100 bcast=192. [root@bigboy tmp]# nmblookup -d 2 ‘*’ added interface ip=192.qxd 2/25/05 10:03 AM Page 183 Testing Samba Client/Server Network Connectivity 183 2.1.168.100 bcast=192. Use the nmblookup -B samba-server-IP-address _SAMBA_ command on the server to determine whether the Samba software is running correctly.1.1.255.1. Use the smbclient //samba-server/tmp command to attempt a commandline login to the Samba server.100 ( 192.1. [root@bigboy tmp]# nmblookup -B 192.1. This should return a message that the login was successful.1. If the test fails.255 nmask=255.103 *<00> 192.255 Got a positive name query response from 192.1.1.13Harrison_ch12. This should return answers from all locally connected clients and servers. This should return the server’s IP address if the software is running correctly.0 querying * on 192.100 __SAMBA__<00> [root@bigboy tmp]# 3.168.168. Use the nmblookup -d 2 “*” command on the server to tell it to broadcast a query message to the network.1. If .168. [root@bigboy tmp]# nmblookup -B 192.168.168.1.100 ( 192.168. use the Linux password of the account with which you logged in.100 ) Got a positive name query response from 192.255 nmask=255.103 “*” querying * on 192.1.255.168. Use the nmblookup -B client-IP-address “*” command on the server to determine whether the client is accepting Samba queries.168.1.100 192.100 *<00> [root@bigboy tmp]# 5.168. You can test other accounts by adding the -U accountname option at the end of the command line. it usually fails if either your client or server has an incorrect subnet mask configured on their NIC cards.255. This should return the client’s IP address if it is running correctly.168.168.100 *<00> 192.1.103 ) Got a positive name query response from 192. Another source of the problem could be that the Client for Microsoft Windows or File and Printer Sharing for Microsoft Networks settings on the client’s NIC haven’t been selected.168.168.255.1.100 __SAMBA__ querying __SAMBA__ on 192.1.100 ) 192.168.1.103 ( 192.168. check to see whether the client is running firewall software that could prevent communication. You also could have entered an incorrect IP address.168.1.168. When prompted for a password.168.1.

(In the example below. This test attempts to log in using the username and password with which you logged into the PC. hosts deny.13Harrison_ch12. Make sure the corresponding Samba user has been created. C:\> . or that the password entered is incorrect. A “Network name not found” message usually points to an incorrect NetBIOS configuration on the client. 4.conf file. If you’re successful you should see: C:\>net view \\bigboy Shared resources at \\bigboy Samba Server Share name Type Used as Comment -----------------------------------------------peter Disk Home Directories The command completed successfully. [root@bigboy tmp]# smbclient //bigboy/TMP Password: Anonymous login successful Domain=[HOMENET] OS=[Unix] Server=[Samba 3. Click on the Internet Properties (TCP/IP) menu option and then click on the Properties button. Click on the Advanced button and then on the WINS tab. 2.2-7.0. Right-click on the network connection and select Properties.FC1] tree connect failed: NT_STATUS_BAD_NETWORK_NAME [root@bigboy tmp]# A message that warns of an invalid or bad network name could mean that the tmp service on the Samba server isn’t correctly configured. Log into the Windows workstation as a Samba user. 3. If it fails. You may also need to add the name of the Samba server to the PC’s C:\WINDOWS\system32\drivers\etc\lmhosts file. and invalid users parameters are set correctly in your smb. press the Enter key when prompted for a password. Click on the Network Connections icon in the Windows Control Panel. the username is peter. and enable Windows name resolution via DNS using the advanced TCP/IP settings menu on the PC.qxd 2/25/05 10:03 AM Page 184 184 Samba Security and Troubleshooting Chapter 12 you are doing this as user root. 6. Add the IP address of the Samba server to the WINS server settings. You can get to this menu using the following method: 1. then make sure your hosts allow.) Use the net view \\samba-server command to log into the Samba server from the command line and get a listing of your shares. Messages related to bad passwords could mean that the user’s account doesn’t exist. that their smbpasswd wasn’t created.

“Windows. C:\> Make sure your password encryption is set correctly in the smb. issue the nmblookup -M domain command to ensure that there is a master browser for the domain. . You should also be able to browse through the shares as well.conf file. You will need to set the encrypt passwords option in the [global] section of the smb. 8.conf file to yes. Once all this has tested positively.conf file hasn’t been configured to automatically use the PC user’s username as the Samba login name. Successful attempts should list the IP address of the master browser server. Try to map a drive letter to the user’s default login directory on the Samba server. As you can see. you should be able to see your domain under Windows’ My Network Places located in file manager or in the Start Menu. Remember that doing so may make logins from Windows 95/98/ME clients fail.1.100 homenet<1d> [root@bigboy home]# This may fail with some Windows NT clients if the Samba server has been configured not to use encrypted passwords. C:\> 7. C:\>net use x: \\bigboy\peter The command completed successfully.conf. This is done with the net use x: \\samba-server\share command.168. If not. Linux.1. Here we want user “peter” to have a DOS drive X: map to Peter’s Linux home directory on the Samba server.255 192. Make sure you have correctly configured the encrypt passwords option in the [global] section of smb. Access is denied. Log into the Windows workstation as a Samba user. You can do this by setting the user=username option in the [tmp] section of the smb. you’ll need to make sure that the preferred master parameter is set to yes in the [global] section of smb. [root@bigboy home]# nmblookup -M homenet querying fedora on 192.conf.qxd 2/25/05 10:03 AM Page 185 Testing Samba Client/Server Network Connectivity 185 If there is no user account. and Samba.” newer versions of Windows send encrypted passwords only. From the Samba server. As stated in Chapter 10. the test will fail and you will see: C:\> net view \\bigboy System error 5 has occurred.conf file.13Harrison_ch12. Failure could also mean that the server’s smb. it is sometimes best to make all your clients run similar versions of the Windows operating system.168.

[global] . BASIC SAMBA SECURITY You can restrict connections to your server on both a per-interface and a pernetwork basis in the [global] section of the smb.0. interfaces = eth0 lo CONCLUSION By now you should have a fully functional Samba-based network that is suitable for the small office or home.168.” discusses how to configure wireless NICs in Linux servers for this very reason. . “Linux Wireless Networking. so I have included it here.168.qxd 2/25/05 10:03 AM Page 186 186 Samba Security and Troubleshooting Chapter 12 CHECKING THE SAMBA LOGS Samba stores all its log files in the /var/log/samba directory. ☞ Specifying the interfaces on which Samba will be active. A wireless network in some cases would be ideal.conf file. the 192. Chapter 13. try searching the nmbd.0/24 127.1.0/24 has been included as a valid network.1.168.0/8 in your configuration. If the network is located in the home. especially if you’re not sure whether there is any bidirectional connectivity between the Samba server and the workstation.0. ☞ Configuring Samba to deny all connections by default and then allowing specified hosts through with the hosts allow and hosts deny settings. bind interfaces only = Yes hosts deny = ALL hosts allow = 192.log and smbd.1.. Interface eth0 is on the 192. This type of security is activated by: ☞ Setting the bind interfaces only parameter to yes.log files for clues. Always remember to include your loopback interface lo and the loopback interface’s network 127.13Harrison_ch12. SAMBA NETWORK TROUBLESHOOTING It is always a good idea to use such network troubleshooting tools as tcpdump to do detailed troubleshooting.. you may want to hide your server where it is less intrusive due to its physical presence or to the noise of its power supply fan or hard drive. If you find yourself having difficulties.0/24 network. In this case. You also can include the IP addresses of individual hosts in this list.

however. but at the time it wasn’t. I made it wireless so I could take it out of my bedroom where the DSL connection entered my apartment. I thought wireless Linux would be easy. Wireless Tools is more convenient to use as it requires fewer configuration steps and the RPM package 187 . Looking back. make sure you purchase a NIC that is Linux compatible. You also need to decide on the wireless Linux package you intend to use: Linux-WLAN or Wireless Tools.qxd 2/25/05 10:04 AM Page 187 C H A P T E R 13 Linux Wireless Networking IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ Wireless Linux-Compatible NICs Common Wireless Networking Terms Networking with Linux Wireless Tools Networking with Linux-WLAN Troubleshooting Your Wireless LAN Wireless Networks in Businesses Conclusion My very first Linux web server was an ancient desktop computer that I bought from a secondhand store that advertised it as being “very obsolete.linuxhomenetworking. Before you consider making a Linux server wireless capable.” It was cheap and it worked. I really did it for the challenge.11 standard have many advantages— not just the aesthetic one I mentioned.14Harrison_ch13. and wireless networks offer relatively easy and low cost deployment with security that’s becoming increasingly better. This is the chapter about what started it all. I had so many headaches with it that I thought one of my very first Web pages should be about my little nightmare warning people about how to do it right. This was how www. and also because we all get stupid some of the time. Wireless networks using the 802.com was born. The hardware is commonly available. Spending more money than I should have on the antique. but it was ugly and noisy—so noisy that it quickly became too loud to tolerate.

They then supply different drivers with each new card to make them work. so do your research before proceeding. which worked with Linux. For that. which is usually connected to a regular external or integrated router for communication to the Internet. Always check the compatibility lists before buying your wireless hardware. but this prevents them from communicating with any other communications path. WIRELESS LINUX-COMPATIBLE NICS Not all wireless NICs work with Linux. check www. the original WMP won’t work without upgrading the firmware. you need a WAP on your network. Even so.linux-wlan.7 (Broadcom chip set) and version 4 (InProComm chip set) do not. The Linksys WMP11 wireless card is a good example of this confusion. Service Set ID. COMMON WIRELESS NETWORKING TERMS Learning the ins and outs of wireless Linux networks will be easier if we’re all speaking the same language. do your homework. For this reason.14Harrison_ch13. Frequently Linux drivers for the new cards are unavailable. You can find hardware compatibility lists for Wireless Tools quite easily on popular search engines. It is possible to buy cards with the same model number from the same vendor with very different circuitry. take time to become familiar with three key wireless terms: wireless access point. .org for the latest list of compatible hardware. analogous to switches in regular wired networks. I will explain all this later. therefore. Before proceeding. and shared encryption key. WAPs are. Learn them now. The original version of the card used the Intersil Prisim chip set. but the newer version 2. Don’t worry. Servers can communicate with one another without a WAP if their NICs are configured in Ad Hoc mode. For Linux-WLAN.qxd 2/25/05 10:04 AM Page 188 188 Linux Wireless Networking Chapter 13 doesn’t have to be reinstalled every time you upgrade your kernel as with Linux-WLAN. Wireless NIC manufacturers are notorious for changing the chip sets on their cards depending on the price of the components. because you’ll see them throughout the chapter. In the most common operating mode (Infrastructure mode). The packages support different NICs. Wireless Access Points A wireless access point (WAP) is a device that acts as the central hub of all wireless data communications. all wireless servers communicate with one another via the WAP.

this makes it easy to eavesdrop on a neighboring network. NETWORKING WITH LINUX WIRELESS TOOLS The Linux Wireless Tools package. You can configure your NIC’s IP settings as if the NIC were a regular Ethernet device. Make sure everything works normally first and then encrypt later. but it will not function correctly as its wireless settings haven’t been configured yet. I’ll stick with ESSID unless the term SSID is relevant to an application. unlike Linux-WLAN. The desired ESSID is then set on both wireless NICs and WAPs. probably meets most of your 802. installed by default. but also to encrypt your wireless data whenever possible. Most wireless software packages enable you to view all the available ESSIDs within range and give you the option of selecting the corresponding wireless LAN (WLAN) to join. if you encrypt your wireless network’s data then you also need to use a shared encryption key on all the NICs and WAPs. and therefore it is best to not only change your ESSID from the factory defaults. Each wireless network needs to be assigned an ESSID that doesn’t match that of any neighboring networks within its range of operation. Encryption Keys Encryption is a method of encoding data so that only people with the secret key to unlock the code can view the original data. The Extended Service Set ID (ESSID) helps prevent the garbling of messages. After you use the ifup command the NIC becomes active. Some software packages allow you to use a plain text key. The term ESSID is frequently interchangeably referred to as a Service Set ID (SSID) by many vendors.11a/b needs. but the more secure ones treat the key like a password and prompt you to enter your chosen key twice without your actually being able to ever see the unencrypted key again.11a/b wireless networks typically found in a home environment share the same frequency range with one another so it is possible for your computer to hear the traffic meant for somebody else’s nearby network.qxd 2/25/05 10:04 AM Page 189 Networking with Linux Wireless Tools 189 Extended Service Set ID The 802. It is always best to encrypt your network data last to simplify troubleshooting. . Its main advantage is that. Using iwconfig for Wireless Tools Configuration After physically installing your Linux-compatible NIC. As expected. which in turn ignore all traffic using other identifiers. you don’t have to reinstall it every time you upgrade your kernel. you need to configure your NIC’s IP and wireless settings before Wireless Tools works. Unfortunately.14Harrison_ch13.

168. forgetting to do so can be problematic. For example.255. you need to make the changes permanent.100 NETMASK=255. however. and finally choose Managed (a WAP on present of the network) or Ad-hoc (no WAP) for the wireless mode. including the SSID and the wireless mode. They are not needed for IP aliases.255.14Harrison_ch13. Your wireless NIC should function as if it were a regular Ethernet NIC using the ifup and ifdown commands. provide the ESSID to use (in this case homenet). . then the commands would be: iwconfig eth0 mode Managed iwconfig eth0 essid homenet Your NIC should now become fully functional.0 ONBOOT=yes BOOTPROTO=static 2.qxd 2/25/05 10:04 AM Page 190 190 Linux Wireless Networking Chapter 13 The most commonly used command in Wireless Tools is iwconfig. # # Wireless configuration # TYPE=Wireless MODE=Managed ESSID=homenet These commands need only be on the main interface file. Managed means that there is a wireless access point (WAP) on the network and Ad-hoc signifies that there is none. The next section shows how to make these iwconfig changes permanent. which you can use to configure most of the wireless parameters. 1. Configure your /etc/sysconfig/network-scripts/ifcfg-eth0 file normally as if it were a regular Ethernet NIC. iwconfig Permanent Wireless Tools Configuration Once your ad-hoc configuration has been completed. For the wireless mode. You will need to run these commands each time you use the ifup command. Add the following statements to the end to specify that the NIC is wireless.1. DHCP Version ============ DEVICE=eth0 USERCTL=yes ONBOOT=yes BOOTPROTO=dhcp Fixed IP Version ================= DEVICE=eth0 IPADDR=192. if your wireless NIC is named eth0 and your managed network’s ESSID is homenet.

. It is generally more difficult to install than Wireless Tools and has fewer troubleshooting tools. There should be ten characters in total: iwconfig eth0 key 967136deac The same rules (no colons or non-hexidecimals between the ten total characters) apply when using the /etc/sysconfig/network-scripts files to add encryption: KEY=967136deac NETWORKING WITH LINUX-WLAN Linux-WLAN is one of the original wireless LAN products developed for Linux. making it a desirable alternative based on the NIC card you have available. You’ll notice that Linux-WLAN uses the term SSID instead of ESSID in its configuration files.qxd 2/25/05 10:04 AM Page 191 Networking with Linux-WLAN 191 Wireless Tools Encryption It is usually best to test your network in an unencrypted state before activating the additional security. This allows you to limit your troubleshooting activities to basic wireless settings.14Harrison_ch13. The advantage of the nwepgen command is that you can provide an easily remembered string that it will consistently encode in to an ESSID key. You can use any one of the rows of characters to create a 40-bit key. which are composed of numeric values between 0 and 9 and alphabetic characters between A and F: [root@bigboy tmp]# /sbin/nwepgen ketchup 64:c1:a1:cc:db 2b:32:ed:37:16 b6:cc:9e:1b:37 d7:0e:51:3f:03 [root@bigboy tmp]# When using iwconfig to add encryption. Encryption requires an encryption key that you can make up yourself or generate with the /sbin/nwepgen command that comes with the kernel-wlan-ng RPM (more on this coming up). but it does have wide ranging hardware support. be sure that there are no colons or any other non-hexadecimal characters between the characters of the key. If you don’t have nwepgen. without the additional complications of encryption. then remember to use hexadecimal numbers.

Linux-WLAN doesn’t identify the wireless NIC as an Ethernet eth device. Chapter 5.nc.raleigh.14Harrison_ch13. “Installing RPM Software. as in kernel-pcmcia-cs-3.us. ☞ Once configured. versions of the driver files at Remember to download the files for the correct kernel type. If you need a refresher. ☞ Always be prepared to check your syslog /var/log/messages file for errors if things don’t work: It is a good source of information. Chapter 6. In Fedora Core. Identifying the Correct RPMs You http://prism2. can find RPM Determine the Kernel Type Use the uname -p command. When searching for the RPMs. It’s a good idea to decide on a common SSID and stick with it. always use the uname version: . but as a wlan device: This is good to know in order to avoid confusion when troubleshooting.31-13. OS version.i386. “Troubleshooting Linux with syslog. and kernel version. the package name is pcmcia-cs and in Red Hat 9 and earlier it is kernel-pcmcia-cs.1. Downloading and installing RPMs isn’t hard. ☞ You may get “device unknown” or “no such device” errors related to the wlan device in the /var/log/messages file if you use older unpatched versions of the Linux-WLAN software: Always use the most recent versions to make the installation smoother. The Linux version may not match the CPU you have installed. “Why Host Your Own Site. The Bigboy server discussed in Chapter 1.qxd 2/25/05 10:04 AM Page 192 192 Linux Wireless Networking Chapter 13 Linux-WLAN Preparation Here are some pointers you’ll need to remember prior to using the LinuxWLAN product: ☞ All devices on a wireless network must use the same Network Identifier or SSID to communicate with each other: The default SSID for Linux-WLAN is linux-wlan. remember that the filename usually starts with the software package name and a version number. ☞ Before installing the Linux-WLAN software for PCMCIA cards such as the Linksys WPC11 you will need to install the RPM packages that support PCMCIA: This step isn’t necessary for such true PCI cards as the Linksys WMP11. the default SSID for your windows NICs may be different.” covers how to do this in detail.” shows you how to set up syslog error logging to be more sensitive to errors.rpm.” is running an i686 version of Linux.unixguru.

remember that the filename usually starts with the software package name by a version number: kernel-wlan-ng-0.and PCI-based NICs.rpm kernel-wlan-ng-pcmcia-0.2.i686.20-8 [root@bigboy tmp]# Installing the RPMs After you have all this Linux information.1-pre14.1-pre14.2. you need to download and install the base. [root@bigboy tmp]# cat /etc/redhat-release Red Hat Linux release 9 (Shrike) [root@bigboy tmp]# [root@zero root]# cat /etc/fedora-release Fedora Core release 1 (Yarrow) [root@zero root]# Determine the Kernel Version You can use the uname -r command to figure out the kernel version.1-pre14.i686.2115-0.i686.rpm kernel-wlan-ng-pci-0. You can also look at the /etc/issue file for other versions of Linux. server Bigboy is running Red Hat version 9.14Harrison_ch13. while Zero is running Fedora Core 1.rpm Note There are different RPMs for PCMCIA.4. module.4. Bigboy is running version 2.2.1-pre14.1.rpm kernel-wlan-ng-modules-fc1.qxd 2/25/05 10:04 AM Page 193 Networking with Linux-WLAN 193 [root@bigboy tmp]# uname -p i686 [root@bigboy tmp]# Determine the OS Version hat-release One of the easiest ways to determine the OS version is to view the /etc/ redor the /etc/fedora-release file.20-8: [root@bigboy tmp]# uname -r 2. When searching for the RPMs. In this case. .2.0.i686. and interface packages. In this case. The base and modules RPMs need to be installed in all cases.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 194

194

Linux Wireless Networking

Chapter 13

Notice the sequence of installation in this sample output. Double-check your preparation steps and the RPM versions if the very last line of the installation gives a result code that is not success.
[root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-0.2.1-pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng ##################################### [100%] [root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-modules-fc1.1.2115-0.2.1pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng-modules-##################################### [100%] [root@bigboy tmp]# [root@bigboy tmp]# rpm -Uvh kernel-wlan-ng-pcmcia-0.2.1-pre14.i686.rpm Preparing... ##################################### [100%] 1:kernel-wlan-ng-pci ##################################### [100%] Adding prism2_pci alias to /etc/modprobe.conf file... ***NOTE*** YOU MUST CHANGE THIS IF YOU HAVE A PLX CARD!!! The default wlan0 network configuration is DHCP. Adjust accordingly. ACHTUNG! ATTENTION! WARNING! YOU MUST configure /etc/wlan/wlan.conf to define your SSID! YOU ALSO must configure /etc/wlan/wlancfg-SSID to match WAP settings! (---> replace SSID in filename with the value of your SSID) If you get an error after this point, there is either a problem with your drivers or you don’t have the hardware installed! If the former, get help! Starting WLAN Devices:message=dot11req_mibset mibattribute=dot11PrivacyInvoked=false resultcode=success message=dot11req_mibset mibattribute=dot11ExcludeUnencrypted=false resultcode=success [root@bigboy tmp]#

Note If you upgrade your Linux kernel, you’ll have to reinstall Linux-WLAN all over again. This will also create new versions of your /etc/sysconfig/networkscripts/ifcfg-wlan0, /etc/wlan/wlan.conf, and /etc/pcmcia/wlan-ng.opts files. You may have to restore these from the automatically saved versions. Linux-WLAN Post Installation Steps After the RPMs are installed, you need to configure the new NIC to be compatible with your network.
wlan0

wireless

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 195

Networking with Linux-WLAN

195

Configure the New wlan0 Interface

Edit /etc/sysconfig/network-scripts/ifcfg-wlan0 to include these new lines:
DHCP Version ============ DEVICE=wlan0 USERCTL=yes ONBOOT=yes BOOTPROTO=dhcp Fixed IP Version ================= DEVICE=wlan0 IPADDR=192.168.1.100 NETMASK=255.255.255.0 ONBOOT=yes BOOTPROTO=static

In the fixed IP version you also need to substitute your selected IP, netmask, network, and broadcast address with those above. Plus, make sure you have the correct gateway statement in your /etc/sysconfig/network file, for example. GATEWAY[eq]192.168.1.1.
Disable Your Existing Ethernet NIC

You may want to disable your existing eth0 Ethernet interface after installing the drivers. Add an ONBOOT=no entry to the /etc/sysconfig/networkscripts/ifcfg-eth0 file. This disables the interface on reboot or when /etc/init.d/network is restarted.
Select the Wireless Mode and SSID

All the configuration files are located in the /etc/wlan directory. The package allows your server to be connected to up to three wireless LANs. You specify the SSIDs (LAN IDs) for each wireless LAN in the /etc/wlan/wlan.conf file. In the example, I make the wlan0 interface join the homenet WLAN, as well as instruct the WLAN driver to scan all wireless channels for SSIDs.
# # Specify all the wlan interfaces on the server # WLAN_DEVICES=”wlan0” # # Specify whether the server should scan the network channels # for valid SSIDs # WLAN_SCAN=y # # Specify expected SSIDs and the wlan0 interface to which it should # be tied # SSID_wlan0=”homenet” ENABLE_wlan0=y

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 196

196

Linux Wireless Networking

Chapter 13

Each WLAN specified in the /etc/wlan/wlan.conf file has its own configuration file. Copy the /etc/wlan/wlancfg-DEFAULT file to a file named /etc/wlan/wlancfg-SSID (replace SSID with the actual SSID for your WAP). This line configures for the homenet SSID:
[root@bigboy wlan]# cp wlancfg-DEFAULT wlancfg-homenet

Start Linux-WLAN Start the wlan process and test for errors in the file /var/log/messages. All the result codes in the status messages should be success. You may receive the following error, however, which the WLAN RPM Web site claims is “harmless.”
Error for wireless request “Set Encode” (8B2A) : SET failed on device wlan0 ; Function not implemented. Error for wireless request “Set ESSID” (8B1A) : SET failed on device wlan0 ; Function not implemented.

With PCI cards, you can restart Linux-WLAN by restarting the WLAN daemon:
[root@bigboy tmp]# service wlan restart [root@bigboy tmp]# ifup wlan0

With PCMCIA cards, you can start Linux-WLAN by restarting the Linux PCMCIA daemon:
[root@bigboy tmp]# service pcmcia restart [root@bigboy tmp]# service network restart

Testing Linux-WLAN Now check to see if the IP address of the wlan0 interface is okay. Refer to the upcoming troubleshooting section if you cannot ping the network’s gateway.
[root@bigboy tmp]# ifconfig -a [root@bigboy tmp]# ping <gateway-address>

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 197

Networking with Linux-WLAN

197

Linux-WLAN Encryption for Security One of the flaws of wireless networking is that all the wireless clients can detect the presence of all available network SSIDs and have the option of joining any of them. With encryption, the client must have a membership encryption password that can also be represented as a series of Wireless Encryption Protocol (WEP) keys. The wlan.conf file (Red Hat 8.0 RPMs), wlan-SSID file (Red Hat 9/Fedora Core 1 RPMs), or /etc/pcmcia/wlan-ng.opts file (PCMCIA-type NICs) is also used to activate this feature. Tip I strongly recommend that you first set up your network without encryption. Only migrate to an encrypted design after you are satisfied that the unencrypted design works correctly. To invoke encryption, you have to set the dot11PrivacyInvoked parameter to true. You also must state which of the keys will be used as the default starting key via the dot11WEPDefaultKeyID parameter. You then have the option of either providing a key-generating string (simple password) or all four of the keys. In the example below, ketchup is the password used to automatically generate the keys.
#=======WEP=========================================== # [Dis/En]able WEP. Settings only matter if PrivacyInvoked is true lnxreq_hostWEPEncrypt=false # true|false lnxreq_hostWEPDecrypt=false # true|false dot11PrivacyInvoked=true dot11WEPDefaultKeyID=1 dot11ExcludeUnencrypted=true # true|false, in AP this means WEP # is required for all STAs # If PRIV_GENSTR is not empty, use PRIV_GENTSTR to generate # keys (just a convenience) PRIV_GENERATOR=/sbin/nwepgen # nwepgen, Neesus compatible PRIV_KEY128=false # keylength to generate PRIV_GENSTR=”ketchup” # or set them explicitly. Set genstr or keys, not both. dot11WEPDefaultKey0= # format: xx:xx:xx:xx:xx or dot11WEPDefaultKey1= # xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx:xx dot11WEPDefaultKey2= # e.g. 01:20:03:40:05 or dot11WEPDefaultKey3= # 01:02:03:04:05:06:07:08:09:0a:0b:0c:0d

Not all devices on your network will use the same algorithm method to generate the encryption keys. You may find the same generator string will not create the same keys, rendering intra-network communication impossible. If this is the case, you can use the /sbin/nwepgen program to generate the keys after you provide an easy to remember key generator string. Once you have the four sets of keys, you’ll have to add them individually and in sequence to

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 198

198

Linux Wireless Networking

Chapter 13

wlan.conf, wlan-SSID or /etc/pcmcia/wlan-ng.opts file and set the PRIV_ GENSTR parameter to a null string of “” (the quotes are important). Here is how you can use nwepgen to create the keys with a generator string of ketchup. [root@bigboy tmp]# /sbin/nwepgen ketchup 64:c1:a1:cc:db 2b:32:ed:37:16 b6:cc:9e:1b:37 d7:0e:51:3f:03 [root@bigboy tmp]#

the

In this case, your wlan.conf or wlan-SSID file would look like this:
PRIV_GENSTR=”” # or set them explicitly. Set genstr or keys, not both. dot11WEPDefaultKey0= 64:c1:a1:cc:db dot11WEPDefaultKey1= 2b:32:ed:37:16 dot11WEPDefaultKey2= b6:cc:9e:1b:37 dot11WEPDefaultKey3= d7:0e:51:3f:03

Remember that all devices on your network, including all wireless NICs and WAPs, need to have the same keys and default key for this to work.
De-activating Encryption

In some cases, NICs without full Linux-WLAN compatibility freeze up after a number of hours of working with encryption. The steps to reverse encryption are: 1. Set the configuration file parameter dot11PrivacyInvoked to false. 2. Stop Linux-WLAN, and disable the wireless wlan0 interface.
[root@bigboy tmp]# service wlan stop Shutting Down WLAN Devices:message=lnxreq_ifstate ifstate=disable resultcode=success [root@bigboy tmp]# ifdown wlan0

3. The driver is still loaded in memory with the old encryption parameters, even though it is not active. Linux frequently loads device driver software, such as those that govern the operation of NICs, as modules that the kernel, or Linux master program, uses in its regular operation. Use the lsmod command to display a list of loaded modules. You’ll be most interested in the modules associated with 802.11 wireless protocols, which appear here as p80211 and prism2_pci:
[root@bigboy tmp]# lsmod Module Size ... ... Used by Not tainted

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 199

Troubleshooting Your Wireless LAN

199

prism2_pci p80211 ... ... [root@bigboy tmp]#

66672 20328

1 1

(autoclean) [prism2_pci]

Sometimes your NIC may use the prism drivers:
[root@bigboy tmp]# lsmod Module Size ... ... orinoco 45517 hermes 6721 ... ... [root@bigboy tmp]#

orinoco

chip set drivers instead of the

Used by

1 orinoco_pci 2 orinoco_pci,orinoco

4. Now that you have identified the driver modules in memory, unload them with the rmmod command:
[root@bigboy tmp]# rmmod prism2_pci [root@bigboy tmp]# rmmod p80211

5. Restart Linux-WLAN, reactivate the functional again:

wlan0

interface, and you should be

[root@bigboy tmp]# service wlan start Starting WLAN Devices:message=lnxreq_hostwep resultcode=no_value decrypt=false encrypt=false [root@bigboy tmp]# ifup wlan0

/var/log/messages

If you fail to reload the driver modules, you’ll get errors in your file and your NIC card will operate in an encrypted mode only:
Jan 2 18:11:12 bigboy kernel: failed,result=-110 Jan 2 18:11:18 bigboy kernel: timeout(1), reg=0x8021. Jan 2 18:11:18 bigboy kernel: failed. Jan 2 18:11:18 bigboy kernel: prism2sta_ifstate: hfa384x_drvr_start() hfa384x_docmd_wait: hfa384x_cmd hfa384x_drvr_start: Initialize command hfa384x_drvr_start: Failed, result=-110

TROUBLESHOOTING YOUR WIRELESS LAN
Linux wireless troubleshooting tools are quite extensive and provide a variety of useful information to help you get your network working. This section covers many important strategies that will complement the use of more conventional procedures such as scanning your /var/log/messages file.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 200

200

Linux Wireless Networking

Chapter 13

Check the NIC Status When using WLAN methodology, the iwconfig, iwlist, and iwspy commands can provide useful information about the status of your wireless network. Take a closer look.
The iwconfig Command

In addition to using the regular ifconfig command to check the status of your NIC, you can use the iwconfig command to view the state of your wireless network, just don’t specify any parameters. Specifically, you can see such important information as the link quality, WAP MAC address, data rate, and encryption keys, which can be helpful in ensuring the parameters across your network are the same. For example:
[root@bigboy tmp]# iwconfig eth0 IEEE 802.11-DS ESSID:”homenet” Nickname:”bigboy” Mode:Managed Frequency:2.462GHz Access Point: 00:09:5B:C9:19:22 Bit Rate:11Mb/s Tx-Power=15 dBm Sensitivity:1/3 Retry min limit:8 RTS thr:off Fragment thr:off Encryption key:98D1-26D5-AC Security mode:restricted Power Management:off Link Quality:36/92 Signal level:-92 dBm Noise level:-148 dBm Rx invalid nwid:0 Rx invalid crypt:2 Rx invalid frag:0 Tx excessive retries:10 Invalid misc:0 Missed beacon:0 [root@bigboy tmp]#

The iwlist Command

The iwlist command can provide further information related to not just the NIC, but the entire network, including the number of available frequency channels, the range of possible data rates, and the signal strength. This example uses the command to verify the encryption key being used by the NIC, which can be very helpful in troubleshooting security related difficulties on your network:
[root@bigboy tmp]# iwlist key ... ... eth0 2 key sizes : 40, 104bits 4 keys available : [1]: 9671-36DE-AC (40 bits) [2]: off [3]: off [4]: off Current Transmit Key: [1] Security mode:open ... ... [root@bigboy tmp]#

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 201

Troubleshooting Your Wireless LAN

201

The iwlist command can verify the speed of the NIC being used, 11Mb/s in this case. This can be helpful in determining possible reasons for network slowness, especially as poor signal quality can result in the NIC negotiating a low bit rate with its WAP.
[root@bigboy tmp]# iwlist rate ... ... eth0 4 available bit-rates : 1Mb/s 2Mb/s 5.5Mb/s 11Mb/s Current Bit Rate:11Mb/s ... ... [root@bigboy tmp]#

For further information on the iwlist command, consult the man pages.
The iwspy Command

The iwspy command provides statistics on the quality of the link between your NIC and another wireless device on the network. It doesn’t run all the time; you have to activate iwspy on your interface first. When not activated, iwspy gives a “no statistics to collect” message:
[root@bigboy root]# iwspy eth0 eth0 No statistics to collect [root@bigboy root]#

To activate the command, specify the target IP address and the wireless NIC interface through which it can be found:
[root@bigboy tmp]# iwspy eth0 192.168.1.1

If you use the iwspy command without the IP address, it provides WLAN statistics with a typical/reference value against which it can be compared. In the example that follows the signal is considered fairly strong, with a 64/92 quality value versus a typical 36/92 value, but it could be weak by the historical values on your network. It’s good to check this from time to time for fluctuations.
[root@bigboy tmp]# iwspy eth0 eth0 Statistics collected: 00:09:5B:C9:19:22 : Quality:0 Signal level:0 Noise level:0 Link/Cell/AP : Quality:64/92 Signal level:-51 dBm Noise level:-149 dBm (updated)

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 202

202

Linux Wireless Networking

Chapter 13

Typical/Reference : Quality:36/92 level:-98 dBm [root@bigboy tmp]#

Signal level:-62 dBm

Noise

To switch off iwspy monitoring, add the off argument:
[root@bigboy root]# iwspy eth0 off

Check for Interrupt Conflicts Devices slotted into your PCI bus are generally assigned an interrupt value by the system, which the system uses to signal its need to communicate with the device. Multiple devices on the bus can have the same interrupt, but the system will access each one using a different memory address to avoid confusion. Sometimes this automatic allocation of interrupt (IRQ) values and memory locations is flawed and overlaps do occur, causing devices to fail. Before configuring your WLAN software, you should ensure that the wireless NIC card doesn’t have an interrupt that clashes with another device in your computer. Insert the card in an empty slot in your Linux box according to the instructions in its manual, reboot, and inspect your /var/log/messages file again:
[root@bigboy tmp]# tail -300 /var/log/messages

Look carefully for any signs that the card is interfering with existing card IRQs. If there is a conflict, there will usually be a warning or “IRQ also used by…” message. If that is the case, move the card to a different slot or otherwise eliminate the conflict by disabling the conflicting device if you don’t really need it. You should also inspect your /proc/interrupts file for multiple devices having the same interrupt:
[root@bigboy tmp]# cat /proc/interrupts 11: 4639 XT-PIC wlan0, eth0 [root@bigboy tmp]# cat /proc/interrupts 11: 4639 XT-PIC wlan0

(potentially bad)

(good)

Interrupt conflicts are usually more problematic with old style PC-AT buses; newer PCI-based systems generally handle conflicts better. The prior (potentially bad) /proc/interrupts example came from a functioning PCIbased Linux box. It worked because, although the interrupt was the same, the base memory addresses that Linux used to communicate with the cards were

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 203

Troubleshooting Your Wireless LAN

203

different. You can check both the interrupts and base memory of your NICs by using the ifconfig -a command:
[root@bigboy tmp]# ifconfig –a eth0 Link encap:Ethernet HWaddr 00:08:C7:10:74:A8 BROADCAST MULTICAST MTU:1500 Metric:1 RX packets:0 errors:0 dropped:0 overruns:0 frame:0 TX packets:0 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:0 (0.0 b) TX bytes:0 (0.0 b) Interrupt:11 Base address:0x1820 wlan0 Link encap:Ethernet HWaddr 00:06:25:09:6A:B5 inet addr:192.168.1.100 Bcast:192.168.1.255 Mask:255.255.255.0 UP BROADCAST RUNNING MULTICAST MTU:1500 Metric:1 RX packets:215233 errors:0 dropped:0 overruns:0 frame:0 TX packets:447594 errors:0 dropped:0 overruns:0 carrier:0 collisions:0 txqueuelen:100 RX bytes:39394014 (37.5 Mb) TX bytes:126738425 (120.8 Mb) Interrupt:11 Memory:c887a000-c887b000 [root@bigboy tmp]#

Kernel Log Errors When you find p80211 Kernel errors in /var/log/messages, they usually point to an incorrectly configured SSID or may also be caused by a NIC card with an outdated firmware version. For example:
Nov 13 22:24:54 bigboy kernel: p80211knetdev_hard_start_xmit: Tx attempt prior to association, frame dropped.

Can’t Ping Default Gateway If you can’t ping the default gateway, first check for kernel log errors. If there are no errors in /var/log/messages and you can’t ping your gateways or obtain an IP address, then check your /etc/sysconfig/networkscripts/ configuration files for a correct IP configuration and your routing table to make sure your routes are OK. You can also check to see if your Linux box is out of range of the WAP using the iwconfig command. Unknown Device Errors Look for “unknown device” or “no such device” errors in your log files or on your screen during installation or configuration. These may be caused by: ☞ An NIC that hasn’t been correctly inserted in the PCI slot. ☞ Incompatible hardware.

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 204

204

Linux Wireless Networking

Chapter 13

For example, you might see incompatible hardware errors in messages:

/var/log/

00:0c.0 Network controller: BROADCOM Corporation: Unknown device 4301 (rev01) Subsystem: Unknown device 1737:4301 Flags: bus master, fast devsel, latency 64, IRQ 5 Memory at f4000000 (32-bit, non-prefetchable) [size=3D8K] Capabilities: [40] Power Management version 2

Or, you might see errors on the screen:
Dec 1 01:28:14 bigboy insmod: /lib/modules/2.4.18-14/net/prism2_pci.o: init_module: No such device Dec 1 01:28:14 bigboy insmod: Hint: insmod errors can be caused by incorrect module parameters, including invalid IO or IRQ parameters. You may find more information in syslog or the output from dmesg Dec 1 01:28:14 bigboy insmod: /lib/modules/2.4.18-14/net/prism2_pci.o: insmod wlan0 failed

A Common Problem with Linux-WLAN and Fedora Core 1 In older versions of Fedora Core 1, the operating system will auto-detect Linux-WLAN-compatible NIC cards and enter a line similar to:
alias eth2 orinoco_pci

in the /etc/modprobe.conf file. In other words, it detects them as an Ethernet eth device instead of a WLAN wlan device. This seems to conflict with the WLAN RPMs, and you’ll get errors like this when starting Linux-WLAN:
Starting WLAN Devices: /etc/init.d/wlan: line 119: Error: Device wlan0 does not seem to be present.: command not found /etc/init.d/wlan: line 120: Make sure you’ve inserted the appropriate: command not found /etc/init.d/wlan: line 121: modules or that your modules.conf file contains: command not found /etc/init.d/wlan: line 122: the appropriate aliase(s).: command not found

You can fix the problem with the proper steps. This example refers to a compatible Orinoco chipset card: 1. Remove the orinoco_pci line from the remove the entry for device wlan0.
/etc/modprobe.conf

file. Do not

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 205

Wireless Networks In Businesses

205

2. Edit your /etc/sysconfig/hwconf file, search for orinoco_pci, and remove the orinoco_pci section that refers to your wireless card. (Each section starts and ends with a single - on a new line.) 3. Reboot. 4. The Linux boot process always runs kudzu, the program that detects new hardware. Kudzu detects the wireless card and asks whether you want to configure it. Choose ignore. This will reinsert the wireless card in the /etc/sysconfig/hwconf file, but not in the /etc/modprobe.conf file. 5. Your NIC should start to function as expected as device wlan0 when you use the ifconfig -a command. Configure the IP address, and activate the NIC as shown earlier in this chapter. The procedure removes all reference to the Orinoco driver in the Linux configuration files and then forces kudzu not to configure the NIC card according to the Linux defaults. The eth device will be recreated, but the ignore option provided to kudzu will prevent the Orinoco entry in the /etc/modprobe.conf from being reinserted, preventing conflict with the LinuxWLAN package’s wlan device.

WIRELESS NETWORKS IN BUSINESSES
Sometimes implementing a wireless network inside a business place becomes necessary. Visiting managers may need a quick connection in a conference room; sales people sharing cubicles my need it as the number of work spaces get exhausted. Perhaps someone is going to set one up on your network anyway—you might as well control this from the beginning. Apart from people who download infected software and e-mail attachments, mobile employees’ notebook computers are usually viewed as a high risk source of unintentional malicious activity as there is even less control over what these employees do than those with fixed workstations. With this in mind, it is usually best to isolate this type of wireless network completely from your internal, trusted, and wired one. Some types of network architectures make the wireless router only have access to the Internet, and no where else, via its own dedicated DSL line. The wireless users then have to use some form of a VPN client to gain access to the office servers just as if they were doing so from home. To reduce the risk of the network being hijacked, be sure to encrypt the traffic and use a proxy server running such software as Squid (see Chapter 32, “Controlling Web Access with Squid”) to limit Internet access to authorized users via some form of pop-up username and password authentication. With this sort of architecture, if the wireless network gets hijacked, your office systems should remain relatively safe. Many WAPs have the option of not advertising their ESSIDs, which prevents users from browsing around to select the nearest available WLAN. Activation of this feature can be inconvenient to users as wireless clients will

14Harrison_ch13.qxd

2/25/05

10:04 AM

Page 206

206

Linux Wireless Networking

Chapter 13

need to know the predefined ESSID to gain LAN access, but more importantly, reduces the risk of an outsider connecting to your wireless LAN by roaming the airwaves for an available WAP. Your WAP may also have the additional security feature of only allowing connections from a list of predefined MAC addresses of all the wireless NICs in your company or department. The more expensive ones have the ability to refer to a database server for this information. There are many other types of wireless methodologies. Please investigate a variety of options before coming to a final conclusion.

CONCLUSION
With the knowledge gained in the chapters in Part 1 of the book you will be able to configure a Linux file and DHCP server on small networks with relative ease. Part 2 will explore the possibility of making your server also become the core of your self-managed dedicated Web site.

15Harrison_Pt2.qxd

2/25/05

10:04 AM

Page 207

P A R T

II

The Linux Web Site Project

207

15Harrison_Pt2.qxd

2/25/05

10:04 AM

Page 208

16Harrison_ch14.qxd

2/25/05

10:06 AM

Page 209

C H A P T E R

14

Linux Firewalls Using iptables

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ What Is iptables? Download and Install the iptables Package How to Start iptables Determining the Status of iptables Packet Processing in iptables Targets and Jumps Important iptables Command Switch Operations Using User-Defined Chains Saving your iptables Scripts Fedora’s iptables Rule Generator Recovering from a Lost Script Loading Kernel Modules Needed by iptables Sample iptables Scripts Troubleshooting iptables Conclusion

Network security is a primary consideration in any decision to host a Web site as the threats are becoming more widespread and persistent every day. One means of providing additional protection is to invest in a firewall. Though prices are always falling, in some cases you may be able to create a comparable unit using the Linux iptables package on an existing server for little or no additional expenditure. This chapter shows how to convert a Linux server into: ☞ A firewall while simultaneously being your home Web site’s mail, Web, and DNS server
209

16Harrison_ch14.qxd

2/25/05

2:37 PM

Page 210

210

Linux Firewalls Using iptables

Chapter 14

☞ A router that will use NAT and port forwarding to both protect your home network and have another Web server on your home network while sharing the public IP address of your firewall Creating an iptables firewall script requires many steps, but with the aid of the sample tutorials, you should be able to complete a configuration relatively quickly.

WHAT IS

IPTABLES?

Originally, the most popular firewall/NAT package running on Linux was ipchains, but it had a number of shortcomings. To rectify this, the Netfilter organization decided to create a new product called iptables, giving it such improvements as: ☞ Better integration with the Linux kernel with the capability of loading iptables-specific kernel modules designed for improved speed and reliability. ☞ Stateful packet inspection. This means that the firewall keeps track of each connection passing through it and in certain cases will view the contents of data flows in an attempt to anticipate the next action of certain protocols. This is an important feature in the support of active FTP and DNS, as well as many other network services. ☞ Filtering packets based on a MAC address and the values of the flags in the TCP header. This is helpful in preventing attacks using malformed packets and in restricting access from locally attached servers to other networks in spite of their IP addresses. ☞ System logging that provides the option of adjusting the level of detail of the reporting. ☞ Better network address translation. ☞ Support for transparent integration with such Web proxy programs as Squid. ☞ A rate limiting feature that helps iptables block some types of denial of service (DoS) attacks. Considered a faster and more secure alternative to ipchains, iptables has become the default firewall package installed under Red Hat and Fedora Linux.

as in iptables-1. Each of these queues is dedicated to a particular type of packet activity and is controlled by an associated packet transformation/ filtering chain. There are three tables in total. Fedora Core will give a simple status message.rpm.” if you need a refresher. remember that the filename usually starts with the software package name by a version number.16Harrison_ch14. HOW TO START IPTABLES You can start.i386.0. stop. [root@bigboy tmp]# PACKET PROCESSING IN IPTABLES All packets inspected by iptables pass through a sequence of built-in tables (queues) for processing.qxd 2/25/05 10:06 AM Page 211 Packet Processing in iptables 211 DOWNLOAD AND INSTALL THE IPTABLES PACKAGE Before you begin. For example: [root@bigboy tmp]# service iptables status Firewall is stopped.91.) When searching for the RPMs. This is hardly used in a home or SOHO environment. you need to make sure that the iptables software RPM is installed. and restart iptables after booting by using the commands: [root@bigboy tmp]# service iptables start [root@bigboy tmp]# service iptables stop [root@bigboy tmp]# service iptables restart To get iptables configured to start at boot. (See Chapter 6. .2. which is responsible for the alteration of quality of service bits in the TCP header. “Installing RPM Software. The first is the mangle table. use the chkconfig command: [root@bigboy tmp]# chkconfig iptables on DETERMINING THE STATUS OF IPTABLES You can determine whether iptables is running or not via the service iptables status command.

. This implies that there was no need to modify the destination IP address of the packet as in prerouting. ☞ INPUT chain: Filters packets destined for the firewall. Table 14. ☞ POSTROUTING chain: NATs packets when the source address of the packet needs to be changed. Address translation occurs after routing. Used with NAT of the source IP address using either one to one or many to one NAT. Filters packets destined to the firewall. The third table is the nat queue.16Harrison_ch14. which is responsible for packet filtering.1 Processing for Packets Routed by the Firewall Queue Function Packet Transformation Chain in Queue FORWARD INPUT OUTPUT NAT Network Address Translation PREROUTING Chain Function Queue Type filter Packet filtering POSTROUTING Filters packets to servers accessible by another NIC on the firewall. Used with NAT of the destination IP address. also known as destination NAT or DNAT. Table 14. It has three built-in chains in which you can place your firewall policy rules: ☞ FORWARD chain: Filters packets to servers protected by the firewall. It has two built-in chains: ☞ PREROUTING chain: NATs packets when the destination address of the packet needs to be changed. ☞ OUTPUT chain: Filters packets originating from the firewall. Filters packets originating from the firewall. Address translation occurs before routing. Facilitates the transformation of the destination IP address to be compatible with the firewall’s routing table.1 provides more details on each queue. which is responsible for network address translation. This is known as source NAT or SNAT.qxd 2/25/05 10:06 AM Page 212 212 Linux Firewalls Using iptables Chapter 14 The second table is the filter queue.

There is an exception: Most rules are related to filtering. POSTROUTING. (Rarely used in SOHO environments. The packet is first examined by your rules in the mangle table’s PREROUTING chain. To help make this clearer.) Modification of the TCP packet quality of service bits before routing occurs. When the destination server decides to reply. the packet undergoes the same sequence of steps. the packet undergoes SNAT before arriving at Network B. INPUT. if necessary. If the packet is destined for a protected network. The rules in the OUTPUT chain of the nat table determine whether address translation is required and the rules in the OUTPUT chain of the filter table are then inspected before the packet is routed back to the Internet. OUTPUT. the firewall needs to reply. At some point. if any.qxd 2/25/05 10:06 AM Page 213 Packet Processing in iptables 213 Queue Type Queue Function Packet Transformation Chain in Queue OUTPUT Chain Function mangle TCP header modification PREROUTING.) You need to specify the table and the chain for each firewall rule you create.1 a TCP packet from the Internet arrives at the firewall’s interface on Network A to create a data connection. if any. This reply is inspected by your rules in the OUTPUT chain of the mangle table. The filter table is therefore the default. FORWARD Network address translation for packets generated by the firewall. then it is filtered by the rules in the INPUT chain of the filter table before being processed by the intended application on the firewall. then it is filtered by the rules in the FORWARD chain of the filter table and. It is then inspected by the rules in the nat table’s PREROUTING chain to see whether the packet requires DNAT. If the packet is destined for the firewall itself. It is then routed.16Harrison_ch14. (Rarely used in SOHO environments. so iptables assumes that any chain that’s defined without an associated table will be a part of the filter table. In Figure 14. . take a look at the way packets are handled by iptables.

1 A diagram of the iptables packet flow.qxd 2/25/05 10:06 AM Page 214 214 Linux Firewalls Using iptables Chapter 14 Packet In NetworkA Routing mangle Table PREROUTING Chain filter Table OUTPUT Chain nat Table PREROUTING Chain nat Table OUTPUT Chain Routing mangle Table OUTPUT Chain Data for the Firewall? No Yes Firewall Reply filter Table FORWARD Chain Local Processing of Data filter Table INPUT Chain nat Table POSTROUTING Chain Packet Out Network B Figure 14. It is now time to discuss the ways in which you add rules to these chains.16Harrison_ch14. .

--to-source <address> [-<address>][:<port>-<port>] SNAT The source IP address is user defined.16Harrison_ch14.qxd 2/25/05 10:06 AM Page 215 Targets and Jumps 215 TARGETS AND JUMPS Each firewall rule inspects each IP packet and then tries to identify it as the target of some sort of operation. Specifies the source IP address and ports to be used by SNAT. N/A --log-prefix “string” Tells iptables to prefix all log messages with a user defined string. The packet is blocked. Used to do source network address translation. Once a target is identified. Qualifiers include icmp-port-unreachable (default) icmp-net-unreachable icmp-host-unreachable icmp-proto-unreachable icmp-net-prohibited icmp-host-prohibited tcp-reset echo-reply DNAT Used to do destination network address translation. The first logs the packet. Table 14. it is common to have two similar rules in sequence. the second drops it.2 lists the built-in targets that iptables uses. The packet information is sent to the syslog daemon for logging. Table 14. iptables stops further processing. --to-destination ipaddress Tells iptables what the destination IP address should be. rewriting the destination IP address of the packet. the packet needs to jump over to it for further processing. Works like the DROP target. continues . but also returns an error message to the host sending the packet that the packet was blocked. rewriting the source IP address of the packet. As you can’t log and drop at the same time. Frequently used to tell why the logged packet was dropped. Most Common Options N/A DROP LOG REJECT The packet is handed over to the end application or the operating system for processing.2 Target ACCEPT Descriptions of the Most Commonly Used Targets Description iptables stops further processing. --reject-with qualifier The qualifier tells what type of reject message is returned. iptables continues processing with the next rule in the table.

Types include icmp. mangle. Deletes all the rules in the selected table.1. iptables Command Switch -t <table> -j <target> -A -F -p <protocol-type> -s <ip-address> -d <ip-address> -i <interface-name> -o <interface-name> In this command switches example: iptables -A INPUT -s 0/0 -i eth0 -d 192. Match protocol.1. The 0/0 representation of an IP address means any. iptables . Flush.qxd 2/25/05 10:06 AM Page 216 216 Linux Firewalls Using iptables Chapter 14 Table 14. The possible built-in tables include: filter.16Harrison_ch14.1 -p TCP -j ACCEPT is being configured to allow the firewall to accept TCP packets coming in on interface eth0 from any IP address destined for the firewall’s IP address of 192.168. nat.2 continued Description Used to do source network address translation. Match source IP address. Append rule to end of a chain. Match destination IP address. such the source IP address and TCP port. tcp. Most Common Options [--to-ports <port>[-<port>]] Target MASQUERADE Specifies the range of source ports to which the original source port can be mapped. Match input interface on which the packet enters.1. Table 14.3 through 14. IMPORTANT IPTABLES COMMAND SWITCH OPERATIONS Each line of an iptables script not only has a jump. By default the source IP address is the same as that used by the firewall’s interface. then the filter table is assumed. Tables 14. Match output interface on which the packet exits.168. but they also have a number of command line options that are used to append rules to chains that match your defined packet characteristics. Jump to the specified target chain when the packet matches the current rule. udp and all.3 General iptables Match Criteria Description If you don’t specify a table.7 list the most common options. There are also options that can be used to just clear a chain so you can start all over again.

The source port is in the range 1024 to 65535 and the destination port is port 80 (www/http). ! --syn means. not a new connection request UDP source port.168.4 Switch Common TCP and UDP Match Criteria Description TCP source port.16Harrison_ch14. can be a single value or a range in the format:starting-port: ending-port -p tcp --sport <port> -p tcp --dport <port> -p tcp --syn -p udp --sport <port> -p udp --dport <port> In this example: iptables -A FORWARD -s 0/0 -i eth0 -d 192.1. You can specify time intervals in the format . Consider another example: iptables -A INPUT -p icmp --icmp-type echo-request \ -m limit --limit 1/s -i eth0 -j ACCEPT The limit feature in iptables specifies the maximum average number of matches to allow per second.58 -o eth1 -p TCP \ --sport 1024:65535 --dport 80 -j ACCEPT iptables is being configured to allow the firewall to accept TCP packets for routing when they enter on interface eth0 from any IP address and are destined for an IP address of 192.58 that is reachable via interface eth1. can be a single value or a range in the format:start-port-number: end-port-number TCP destination port.168. Common ICMP (Ping) Match Criteria Description The most commonly used types are echo-reply and echo-request Table 14.1.5 Matches used with ---icmp-type --icmp-type <type> In this example: iptables -A OUTPUT -p icmp --icmp-type echo-request -j ACCEPT iptables -A INPUT -p icmp --icmp-type echo-reply -j ACCEPT iptables is being configured to allow the firewall to send ICMP echo-requests (pings) and in turn. can be a single value or a range in the format: starting-port: ending-port Used to identify a new connection request. accept the expected ICMP echo-replies. Can be a single value or a range in the format: starting-port: ending-port UDP destination port.qxd 2/25/05 10:06 AM Page 217 Important iptables Command Switch Operations 217 Table 14.

Unlike when -m isn’t used. port> -m --state <state> This is an expansion on the previous example: iptables -A FORWARD -s 0/0 -i eth0 -d 192.6 Switch -m --sport <port. Table 14. A variety of TCP/UDP destination ports separated by commas. they do not have to be within a range. port> Common Extended Match Criteria Description A variety of TCP/UDP source ports separated by commas.168. Source and destination ports are assumed to be the same and they do not have to be within a range.1. This is a common feature of such protocols as an FTP data transfer.443 -j ACCEPT iptables -A FORWARD -d 0/0 -o eth0 -s 192. INVALID: The packet couldn’t be identified. NEW: The packet is the start of a new connection. Unlike when -m isn’t used.qxd 2/25/05 10:06 AM Page 218 218 Linux Firewalls Using iptables Chapter 14 /second. or /day. you can use abbreviations so that 3/second is In this example. /hour.16Harrison_ch14. or an ICMP error. Could be due to insufficient system resources or ICMP errors that don’t match an existing data flow. this feature allows you to filter unusually high volumes of traffic that characterize denial of service (DOS) attacks and Internet worms. /minute.1. ICMP echo requests are restricted to no more than one per second. -m --dport <port. When tuned correctly. port> -m --ports <port. they do not have to be within a range. Here a defense for SYN flood attacks was created by limiting the acceptance of TCP segments with the SYN bit set to no more than five per second. RELATED: The packet is starting a new secondary connection. A variety of TCP/UDP ports separated by commas.168. or the same as 3/s.58 -i eth1 -p TCP \ -m state --state ESTABLISHED -j ACCEPT . iptables -A INPUT -p tcp --syn -m limit --limit 5/s -i eth0 -j ACCEPT You can expand on the limit feature of iptables to reduce your vulnerability to certain types of denial of service attacks.58 -o eth1 -p TCP \ --sport 1024:65535 -m multiport --dport 80. The most frequently tested states are: ESTABLISHED: The packet is part of a connection that has seen packets in both directions.

thereby shortening the total length of all chains the packet has to pass through. Instead of stating the source and destination ports.110. In other words. instead of using a single.1.229. built-in chain for all protocols. protocol-specific chain in the filter table. you can simply allow packets related to established connections using the -m state and --state ESTABLISHED options. The source port is in the range 1024 to 65535 and the destination ports are port 80 (www/http) and 443 (https). For example. For example: iptables -A INPUT -i eth0 -d 206. Table 14.8 summarizes the function of each.168.110. Table 14. The return packets from 192. This feature is frequently used to help streamline the processing of packets. you can use the chain to determine the protocol type for the packet and then hand off the actual final processing to a user-defined. USING USER-DEFINED CHAINS As you may remember.1.2 -j fast-output-queue iptables -A fast-input-queue -p icmp -j icmp-queue-in iptables -A fast-output-queue -p icmp -j icmp-queue-out iptables -A icmp-queue-out -p icmp --icmp-type echo-request \ -m state --state NEW -j ACCEPT iptables -A icmp-queue-in -p icmp --icmp-type echo-reply -j ACCEPT Here six queues help assist in improving processing speed.58 are allowed to be accepted too.58 that is reachable via interface eth1. you can configure iptables to have user-defined chains.2 -j fast-input-queue iptables -A OUTPUT -o eth0 -s 206. you can replace a long chain with a stubby main chain pointing to multiple stubby chains.qxd 2/25/05 10:06 AM Page 219 Using User-Defined Chains 219 Here iptables is being configured to allow the firewall to accept TCP packets to be routed when they enter on interface eth0 from any IP address destined for IP address of 192.8 Chain INPUT OUTPUT fast-input-queue fast-output-queue icmp-queue-out icmp-queue-in Custom Queues Example Listing Description The regular built-in INPUT chain in iptables The regular built-in OUTPUT chain in iptables Input chain dedicated to identifying specific protocols and shunting the packets to protocol specific chains Output chain dedicated to identifying specific protocols and shunting the packets to protocol specific chains Output queue dedicated to ICMP Input queue dedicated to ICMP .168.16Harrison_ch14.229.

The format of the /etc/sysconfig/iptables file is slightly different from that of the scripts shown in this chapter.ESTABLISHED -j ACCEPT -A RH-Firewall-1-INPUT -p tcp -m state --state NEW -m tcp --dport 22 j ACCEPT -A RH-Firewall-1-INPUT -j REJECT --reject-with icmp-host-prohibited COMMIT # Completed on Mon Nov 8 11:00:07 2004 [root@bigboy tmp]# It is never a good idea to edit this script directly because it is always overwritten by the save command and it doesn’t save any comments at all.[0:0] -A INPUT -j RH-Firewall-1-INPUT -A FORWARD -j RH-Firewall-1-INPUT -A RH-Firewall-1-INPUT -i lo -j ACCEPT -A RH-Firewall-1-INPUT -p icmp -m icmp --icmp-type 255 -j ACCEPT -A RH-Firewall-1-INPUT -p esp -j ACCEPT -A RH-Firewall-1-INPUT -p ah -j ACCEPT -A RH-Firewall-1-INPUT -m state --state RELATED. The initialization of built-in chains is automatic and the string “iptables” is omitted from the rule statements.2.16Harrison_ch14. When the system reboots. It is a good place for beginners to start on a test system so that they can see a general rule structure. For these reasons. lokkit saves the firewall rules in a new /etc/sysconfig/iptables file for use on the next reboot. FEDORA’S IPTABLES RULE GENERATOR Fedora comes with a program called lokkit that you can use to generate a very rudimentary firewall rule set. . the iptablesrestore program reads the configuration and makes it the active configuration. Like the service iptables save command.9 on Mon Nov 8 11:00:07 2004 *filter :INPUT ACCEPT [0:0] :FORWARD ACCEPT [0:0] :OUTPUT ACCEPT [144:12748] :RH-Firewall-1-INPUT . Here is a sample /etc/sysconfig/iptables configuration that allows ICMP. IPSec (ESP and AH packets). already established connections. you’re better off writing and applying a customized script and then using the service iptables save command to make the changes permanent. and inbound SSH: [root@bigboy tmp]# cat /etc/sysconfig/iptables # Generated by iptables-save v1.qxd 2/25/05 10:06 AM Page 220 220 Linux Firewalls Using iptables Chapter 14 SAVING YOUR IPTABLES SCRIPTS The service iptables save command permanently saves the iptables configuration in the /etc/sysconfig/iptables file. It prompts for the level of security and then gives you the option of doing simple customizations. which can also make it extremely difficult to follow.

2. then you can edit the output and reload the updated rules when they meet your new criteria with the iptables-restore command. which actually saves a permanent copy of the firewall’s active configuration in the /etc/sysconfig/iptables file. In these situations. Unlike the service iptables save command. it’s best to write scripts that you can comment and then save them to /etc/sysconfig/ iptables.16Harrison_ch14.9 on Mon Nov 8 11:00:07 2004 *filter :INPUT ACCEPT [0:0] :FORWARD ACCEPT [0:0] :OUTPUT ACCEPT [144:12748] :RH-Firewall-1-INPUT . you can use the iptables-save and iptables-restore commands to assist you with the continued management of the server.qxd 2/25/05 10:06 AM Page 221 Recovering from a Lost Script 221 Once you have become familiar with the iptables syntax. This example exports the iptables-save output to a text file named firewall-config: [root@bigboy tmp]# iptables-save > firewall-config [root@bigboy tmp]# cat firewall-config # Generated by iptables-save v1. It makes them much more manageable and readable.ESTABLISHED -j ACCEPT -A RH-Firewall-1-INPUT -p tcp -m state --state NEW -m tcp --dport 22 j ACCEPT -A RH-Firewall-1-INPUT -j REJECT --reject-with icmp-host-prohibited COMMIT # Completed on Mon Nov 8 11:00:07 2004 [root@bigboy tmp]# After editing the firewall-config file with the commands you need. or you might inherit a new system from an administer and cannot find the original script used to protect it. If you redirect the iptables-save screen output to a file with the > symbol. iptables-save displays the active configuration to the screen in /etc/sysconfig/iptables format. RECOVERING FROM A LOST SCRIPT Sometimes the script you created to generate iptables rules may get corrupted or lost.[0:0] -A INPUT -j RH-Firewall-1-INPUT -A FORWARD -j RH-Firewall-1-INPUT -A RH-Firewall-1-INPUT -i lo -j ACCEPT -A RH-Firewall-1-INPUT -p icmp -m icmp --icmp-type 255 -j ACCEPT -A RH-Firewall-1-INPUT -p esp -j ACCEPT -A RH-Firewall-1-INPUT -p ah -j ACCEPT -A RH-Firewall-1-INPUT -m state --state RELATED. you can reload it into the active firewall rule set with the iptables-restore command: [root@bigboy tmp]# iptables-restore < firewall-config .

qxd 2/25/05 10:06 AM Page 222 222 Linux Firewalls Using iptables Chapter 14 Finally. you can eventually convert this firewall-config file into a regular iptables script so that it becomes more easily recognizable and manageable.local # Module to track the state of connections modprobe ip_conntrack # Load the iptables active FTP module. Whenever any type of NAT is required. which tracks TCP connection states. The ip_conntrack_ftp module needs to be added for FTP support and should always be loaded with the ip_conntrack module. the /etc/sysconfig/iptables file doesn’t support the loading of modules. the ip_conntrack module will be needed in any case. The ip_nat_ftp module also needs to be loaded for FTP servers behind a NAT firewall. Pay special attention to the logging example at the end. so you’ll have to add the statements to your /etc/rc. which is run at the end of every reboot.local file. requires ip_conntrack modprobe ip_conntrack_ftp # Load iptables NAT module when required modprobe iptable_nat # Module required for active an FTP server using NAT modprobe ip_nat_ftp SAMPLE IPTABLES SCRIPTS This section provides some sample scripts you can use to get iptables working for you. LOADING KERNEL MODULES NEEDED BY IPTABLES The iptables application requires you to load certain kernel modules to activate some of its functions. As most scripts probably will keep track of connection states. . Unfortunately. the iptable_nat module needs to be loaded.local file: # File: /etc/rc. you should permanently save the active configuration so that it will be loaded automatically when the system reboots: [root@bigboy tmp]# service iptables save If desired.16Harrison_ch14. The script samples in this chapter include these statements only as a reminder to place them in the /etc/rc.

not another.” Here is a sample configuration: # File: /etc/sysctl. It should be a good guide to get you started. Helps to maintain state # Also protects against IP spoofing #--------------------------------------------------------------net/ipv4/conf/all/rp_filter = 1 #--------------------------------------------------------------# Enable logging of packets with malformed IP addresses #--------------------------------------------------------------net/ipv4/conf/all/log_martians = 1 . Respond to queries out # the same interface. UDP. Scripts. For example.conf file. Using of /etc/sysctl.conf #--------------------------------------------------------------# Disable routing triangulation.conf to modify kernel parameters is explained in more detail in Appendix I.16Harrison_ch14. the Linux operating system has a number of built-in protection mechanisms that you should activate by modifying the system kernel parameters in the /proc filesystem via the /etc/sysctl.qxd 2/25/05 10:06 AM Page 223 Sample iptables Scripts 223 The basic initialization script snippet should also be included in all your scripts to ensure the correct initialization of your chains should you decide to restart your script after startup. Note Once you feel more confident. “Codes. and ICMP packets ☞ Provide outbound passive FTP access from the firewall There are also simpler code snippets in Appendix II for Inbound and Outbound FTP connections to and from your firewall. and Configurations.” to find detailed scripts. mail. “Miscellaneous Linux Topics. The appendix shows you how to allow your firewall to: ☞ Be used as a Linux Web. and DNS server ☞ Be the NAT router for your home network ☞ Prevent various types of attacks using corrupted TCP. Basic Operating System Defense You can do several things before employing your firewall script to improve the resilience of your firewall to attack. you can use Appendix II. Other snippets will help you get basic functionality.

.accept_redirects = 0 net.conf.ipv4.all. Required if your firewall is protecting a # network..ipv4..conf.conf. but changes won’t become active in the current boot session until you run the sysctl -p command: [root@bigboy tmp]# sysctl -p . .rp_filter = 1 net.ipv4. NAT included #--------------------------------------------------------------net/ipv4/ip_forward = 1 This configuration will become active after the next reboot.icmp_echo_ignore_broadcasts = 1 [root@bigboy tmp]# .ipv4.16Harrison_ch14.all..accept_source_route = 0 net.all.conf.ipv4.conf.ipv4.send_redirects = 0 net. net.ipv4.qxd 2/25/05 10:06 AM Page 224 224 Linux Firewalls Using iptables Chapter 14 #--------------------------------------------------------------# Disable redirects #--------------------------------------------------------------net/ipv4/conf/all/send_redirects = 0 #--------------------------------------------------------------# Disable source routed packets #--------------------------------------------------------------net/ipv4/conf/all/accept_source_route = 0 #--------------------------------------------------------------# Disable acceptance of ICMP redirects #--------------------------------------------------------------net/ipv4/conf/all/accept_redirects = 0 #--------------------------------------------------------------# Turn on protection from Denial of Service (DOS) attacks #--------------------------------------------------------------net/ipv4/tcp_syncookies = 1 #--------------------------------------------------------------# Disable responding to ping broadcasts #--------------------------------------------------------------net/ipv4/icmp_echo_ignore_broadcasts = 1 #--------------------------------------------------------------# Enable IP routing.tcp_syncookies = 1 net.all.log_martians = 1 net.all.

then they will not reach the INPUT.You may need # them later # # Note: It is best to use the /etc/rc. Included only as a reminder. however. the user defined # chains should be deleted. to initialize your chain and table settings with known values. in any iptables script you write.local example in this # chapter.16Harrison_ch14. This value will not be retained in the # /etc/sysconfig/iptables file. We’ll recreate them in the next step #--------------------------------------------------------------iptables --delete-chain iptables -t nat --delete-chain iptables -t mangle --delete-chain #--------------------------------------------------------------# If a packet doesn’t match one of the built in chains. and if all packets that don’t match the nat and mangle table rules are dropped. and OUTPUT chains for processing. #--------------------------------------------------------------# Load modules for FTP connection tracking and NAT . It is not good policy. and OUTPUT chains should drop packets by default for the best security. Additional ALLOW rules should be added to the end of this script snippet.qxd 2/25/05 10:06 AM Page 225 Sample iptables Scripts 225 Basic iptables Initialization It is a good policy. These tables are queried before the filter table. #--------------------------------------------------------------modprobe modprobe modprobe modprobe ip_conntrack ip_nat_ftp ip_conntrack_ftp iptable_nat #--------------------------------------------------------------# Initialize all the chains by removing all the rules # tied to them #--------------------------------------------------------------iptables –flush iptables -t nat –flush iptables -t mangle –flush #--------------------------------------------------------------# Now that the chains have been initialized. FORWARD. then # The policy should be to drop it #--------------------------------------------------------------iptables iptables iptables iptables iptables --policy INPUT -j DROP --policy OUTPUT -j DROP --policy FORWARD -j DROP -t nat --policy POSTROUTING ACCEPT -t nat --policy PREROUTING ACCEPT #--------------------------------------------------------------# The loopback interface should accept all traffic . The filter table’s INPUT. FORWARD. to make your nat and mangle tables drop packets by default.

The sample script snippet below outlines how to do this. #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#= # # Define networks: NOTE!! You may want to put these “EXTERNAL” # definitions at the top of your script. # #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#= EXTERNAL_INT=”eth0” EXTERNAL_IP=”97.16Harrison_ch14. including checks for Internet traffic from RFC1918 private addresses.25” # External Internet interface # Internet Interface IP address #———————————————————————————————# Initialize our user-defined chains #———————————————————————————————iptables -N valid-src iptables -N valid-dst #———————————————————————————————# Verify valid source and destination addresses for all packets #———————————————————————————————iptables iptables iptables iptables -A -A -A -A INPUT FORWARD OUTPUT FORWARD -i -i -o -o $EXTERNAL_INT $EXTERNAL_INT $EXTERNAL_INT $EXTERNAL_INT -j -j -j -j valid-src valid-src valid-dst valid-dst #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=# # # Source and Destination Address Sanity Checks # # Drop packets from networks covered in RFC 1918 (private nets) # Drop packets from external interface IP # .253. This removes the need to repeat the same statements over and over again.qxd 2/25/05 10:06 AM Page 226 226 Linux Firewalls Using iptables Chapter 14 # Necessary for X-Windows and other socket based services #--------------------------------------------------------------iptables -A INPUT -i lo -j ACCEPT iptables -A OUTPUT -o lo -j ACCEPT Advanced iptables Initialization You may also want to add some more advanced initialization steps to your script.158. which are beyond the scope of this book. More complex initializations would include checks for attacks using invalid TCP flags and directed broadcasts. The complete firewall script in Appendix II outlines most of them. You can take even more precautions to further protect your network. The script also uses multiple user-defined chains to make the script shorter and faster as the chains can be repeatedly accessed.

0.0. but because of Fedora Linux’s yum RPM updater.0/16 $224.Interface eth0 is the internet interface # # Zone transfers use TCP and not UDP.0. .0.16.0.0/8 $172. The following statements will apply not only for firewalls acting as DNS clients but also for firewalls working in a caching or regular DNS server role. #--------------------------------------------------------------# Allow outbound DNS queries from the FW and the replies too # # . Connections initiated by users logged into the Web server will be denied as outbound NEW connection packets aren’t allowed.0. This is not because it is required for the basic functionality of the firewall.0. which will help to keep the server up to date with the latest security patches.0.0/8 255.0.0/12 $192.255.255.0/5 $127. It isn’t necessary to specify these ports for the return leg as outbound packets for all established connections are allowed.0. Most home networks # / websites using a single DNS server won’t require TCP statements # #--------------------------------------------------------------iptables -A OUTPUT -p udp -o eth0 --dport 53 --sport 1024:65535 \ -j ACCEPT iptables -A INPUT -p udp -i eth0 --sport 53 --dport 1024:65535 \ -j ACCEPT Allowing WWW and SSH Access to Your Firewall This sample snippet is for a firewall that doubles as a Web server that is managed remotely by its system administrator via secure shell (SSH) sessions. Inbound packets destined for ports 80 and 22 are allowed.0/8 0. thereby making the first steps in establishing a connection.0.16Harrison_ch14.0/4 -j -j -j -j -j -j -j -j -j -j -j DROP DROP DROP DROP DROP DROP DROP DROP DROP DROP DROP Allowing DNS Access to Your Firewall You’ll almost certainly want your firewall to make DNS queries to the Internet.0.0/4 $240.0.0/16 $EXTERNAL_IP $224.0.255 169.168.qxd 2/25/05 10:06 AM Page 227 Sample iptables Scripts 227 #=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=#=# iptables iptables iptables iptables iptables iptables iptables iptables iptables iptables iptables -A -A -A -A -A -A -A -A -A -A -A valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-src valid-dst -s -s -s -s -s -s -s -d -s -s -d $10.254.0.

then remove the line: -m multiport --dport 80.443 -m multiport \ --sport 1024:65535 #--------------------------------------------------------------# Allow previously established connections # .RELATED \ -j ACCEPT #--------------------------------------------------------------# Allow port 80 (www) and 22 (SSH) connections to the firewall #--------------------------------------------------------------iptables -A INPUT -p tcp -i eth0 --dport 22 --sport 1024:65535 \ -m state --state NEW -j ACCEPT iptables -A INPUT -p tcp -i eth0 --dport 80 --sport 1024:65535 \ -m state --state NEW -j ACCEPT Allowing Your Firewall to Access the Internet This iptables script enables a user on the firewall to use a Web browser to surf the Internet.Interface eth0 is the internet interface #--------------------------------------------------------------iptables -A INPUT -j ACCEPT -m state --state ESTABLISHED.RELATED -i eth0 -p tcp \ If you want all TCP traffic originating from the firewall to be accepted.16Harrison_ch14. as well as by Red Hat Linux servers running up2date.Interface eth0 is the internet interface #--------------------------------------------------------------iptables -A OUTPUT -o eth0 -m state --state ESTABLISHED.443 --sport 1024:65535 . and HTTPS uses port 443. HTTP traffic uses TCP port 80. #--------------------------------------------------------------# Allow port 80 (www) and 443 (https) connections from the firewall #--------------------------------------------------------------iptables -A OUTPUT -j ACCEPT -m state --state NEW \ -o eth0 -p tcp -m multiport --dport 80. Note HTTPS (secure HTTP) is used for credit card transactions frequently. FTP and HTTP are frequently used with yum.qxd 2/25/05 2:38 PM Page 228 228 Linux Firewalls Using iptables Chapter 14 #--------------------------------------------------------------# Allow previously established connections # .

Masquerading also depends on the Linux operating system being configured to support routing between the Internet and private network interfaces of the firewall. using the POSTROUTING and not the MASQUERADE statement.16Harrison_ch14.1. More specifically.1.168. packets related to NEW and ESTABLISHED connections will be allowed outbound to the Internet. This is done by enabling IP forwarding or routing by giving the file /proc/sys/net/ipv4/ip_forward the value 1 as opposed to the default disabled value of 0. The advantage of this is that you never have to specify the NAT IP address. use the FORWARD chain of the filter table.qxd 2/25/05 10:06 AM Page 229 Sample iptables Scripts 229 Allow Your Home Network to Access the Firewall In this example.Interface eth1 is the private network interface #--------------------------------------------------------------iptables -A INPUT iptables -A OUTPUT -j ACCEPT -p all -s 192.168.0 network. types of connections. All traffic between this network and the firewall is simplistically assumed to be trusted and allowed. Keep in mind that iptables requires the iptables_nat module to be loaded with the modprobe command for the masquerade feature to work. Note The masquerade IP address always defaults to the IP address of the firewall’s main interface. traffic from all devices on one or more protected networks will appear as if it originated from a single IP address on the Internet side of the firewall. #--------------------------------------------------------------# Allow all bidirectional traffic from your firewall to the # protected network # .0/24 -o eth1 Masquerading (Many to One NAT) As explained in Chapter 2. In other words.0/24 -i eth1 -j ACCEPT -p all -d 192.1. eth1 is directly connected to a home network using IP addresses from the 192. and possibly even remote servers to have access to your firewall and home network.168. but only packets related to . “Introduction to Networking. An example of this can be seen in the static NAT section that follows. This makes it much easier to configure iptables NAT with DHCP. Further rules will be needed for the interface connected to the Internet to allow only specific ports. you will have to configure iptables to allow packets to flow between the two interfaces. You can configure many to one NAT to an IP alias. To do this.” masquerading is another name for what many call many to one NAT. Once masquerading has been achieved using the POSTROUTING chain of the nat table.

1.Interface eth1 is the private network interface #--------------------------------------------------------------iptables -A POSTROUTING -t nat -o eth0 -s 192.conf example in this # chapter. If a Linux firewall is also your interface to the Internet and you . Included only as a reminder. #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# Allow masquerading # .Interface eth0 is the internet interface # . This value will not be retained in the # /etc/sysconfig/iptables file. the packets are routed via the filter # table’s FORWARD chain.16Harrison_ch14.qxd 2/25/05 10:06 AM Page 230 230 Linux Firewalls Using iptables Chapter 14 ESTABLISHED connections will be allowed inbound. #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc. Port Forwarding Type NAT (DHCP DSL) In many cases home users may get a single DHCP public IP address from their ISPs.RELATED -j ACCEPT iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED. # Allowed outbound: New.ESTABLISHED. This value will not be retained in the # /etc/sysconfig/iptables file.0/24 -d 0/0 \ -j MASQUERADE #--------------------------------------------------------------# Prior to masquerading.RELATED -j ACCEPT Note If you configure your firewall to do masquerading.local example in this # chapter. then it should be used as the default gateway for all your servers on the network. This helps to protect the home network from anyone trying to initiate connections from the Internet. established and related connections # Allowed inbound : Established and related connections #--------------------------------------------------------------iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW. #--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl. Included only as a reminder.168.

qxd 2/25/05 10:06 AM Page 231 Sample iptables Scripts 231 want to host a Web site on one of the NAT protected home servers. Included only as a reminder. the remote server’s IP address.*://’`” #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl. Included only as a reminder.Interface eth0 is the internet interface . #--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Get the IP address of the Internet interface eth0 (linux only) # # You’ll have to use a different expression to get the IP address # for other operating systems which have a different ifconfig output # or enter the IP address manually in the PREROUTING statement # # This is best when your firewall gets its IP address using DHCP. the iptables_nat module has to be loaded and routing has to be enabled for port forwarding to work. This value will not be retained in the # /etc/sysconfig/iptables file.conf example in this # chapter. Routing too must be allowed in iptables with the FORWARD chain.1.local example in this # chapter. then you will have to use port forwarding. this includes all NEW inbound connections from the Internet matching the port forwarding port plus all future packets related to the ESTABLISHED connection in both directions. Here the combination of the firewall’s single IP address. # The external IP address could just be hard coded (“typed in # normally”) #--------------------------------------------------------------external_int=”eth0” external_ip=”`ifconfig $external_int | grep ‘inet addr’ | \ awk ‘{print $2}’ | sed -e ‘s/. #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# Allow port forwarding for traffic destined to port 80 of the # firewall’s IP address to be forwarded to port 8080 on server # 192. As in masquerading.16Harrison_ch14.200 # # . All traffic that matches a particular combination of these factors may then be forwarded to a single server on the private network. Port forwarding is handled by the PREROUTING chain of the nat table. This value will not be retained in the # /etc/sysconfig/iptables file. #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc. and the source/destination port of the traffic can be used to uniquely identify a traffic flow.168.

# Connections on port 80 to the target machine on the private # network must be allowed.200:8080 #--------------------------------------------------------------# After DNAT. not just to a particular port. This is different from masquerading. use SNAT to specify the alias IP address to be used for connections initiated by all other servers in the protected network. and 22 are allowed through by the FORWARD chain. I can’t recommend MASQUERADE. Instead. ☞ Creates a many to one NAT for the 192.158.1.158.ESTABLISHED.29.RELATED -j ACCEPT iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED. all traffic to a particular public IP address. the firewall: ☞ Uses one to one NAT to make the server 192. #--------------------------------------------------------------iptables -A FORWARD -p tcp -i eth0 -o eth1 -d 192.168. Because the firewall has more than one IP address. In this example. only connections on ports 80.168.1. 443.1.168.RELATED -j ACCEPT Static NAT In this example.253.168.168.1. is translated to a single server on the protected subnet. .qxd 2/25/05 10:06 AM Page 232 232 Linux Firewalls Using iptables Chapter 14 # . it will force masquerading as the IP address of the primary interface and not as any of the alias IP addresses the firewall may have. Note Although the nat table translates all traffic to the target servers (192.16Harrison_ch14.100 to 102).100 on your home network appear on the Internet as IP addresses 97.26. Also notice how you have to specify a separate -m multiport option whenever you need to match multiple nonsequential ports for both source and destination.200 \ --dport 8080 --sport 1024:65535 -m state --state NEW -j ACCEPT iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW.253.0 home network in which all the servers appear on the Internet as IP address 97.Interface eth1 is the private network interface #--------------------------------------------------------------iptables -t nat -A PREROUTING -p tcp -i eth0 -d $external_ip \ --dport 80 --sport 1024:65535 -j DNAT --to 192. the packets are routed via the filter table’s # FORWARD chain.1.

26 -i eth0 \ -j DNAT --to-destination 192.26 (1:1 NAT .Inbound) # Anywhere 192.158.1.local example in this # chapter.168.Interface eth0 is the internet interface # .29 (FW IP) # # SNAT is used to NAT all other outbound connections initiated # from the protected network to appear to come from # IP address 97.100 97. Frequently used to NAT connections from # your home network to the Internet # # PREROUTING: # NATs destination IP addresses. Included only as a reminder.253. Frequently used to NAT # connections from the Internet to your home network # # . Included only as a reminder.Outbound) # Anywhere 192.158.1.1.29 # # POSTROUTING: # NATs source IP addresses.158. #--------------------------------------------------------------modprobe iptable_nat #--------------------------------------------------------------# Enable routing by modifying the ip_forward /proc filesystem file # # Note: It is best to use the /etc/sysctl.16Harrison_ch14.158.100 (1:1 NAT . #--------------------------------------------------------------# Load the NAT module # # Note: It is best to use the /etc/rc.conf example in this # chapter.253.253.0/24 97. This value will not be retained in the # /etc/sysconfig/iptables file.253.100 # POSTROUTING statements for 1:1 NAT . #--------------------------------------------------------------echo 1 > /proc/sys/net/ipv4/ip_forward #--------------------------------------------------------------# NAT ALL traffic: ########### # REMEMBER to create aliases for all the internet IP addresses below ########### # # TO: FROM: MAP TO SERVER: # 97.168.168.1.26 Anywhere 192.Interface eth1 is the private network interface #--------------------------------------------------------------# PREROUTING statements for 1:1 NAT # (Connections originating from the Internet) iptables -t nat -A PREROUTING -d 97.qxd 2/25/05 10:06 AM Page 233 Sample iptables Scripts 233 You will have to create alias IP addresses for each of these Internet IPs for one to one NAT to work.158. This value will not be retained in the # /etc/sysconfig/iptables file.253.168.

ESTABLISHED. Checking the Firewall Logs You track packets passing through the iptables list of rules using the LOG target.29 # Allow forwarding to each of the servers configured for 1:1 NAT # (For connections originating from the Internet.443.253.168.158.100 -o eth0 \ -j SNAT --to-source 97.RELATED -j ACCEPT TROUBLESHOOTING IPTABLES A number of tools are at your disposal for troubleshooting iptables firewall scripts.100 \ -m multiport --dport 80. Notice how you # use the real IP addresses here) iptables -A FORWARD -p tcp -i eth0 -o eth1 -d 192.22 \ -m state --state NEW -j ACCEPT # Allow forwarding for all New and Established SNAT connections # originating on the home network AND already established # DNAT connections iptables -A FORWARD -t filter -o eth0 -m state \ --state NEW.1. ☞ Automatically writes an entry to the /var/log/messages file and then executes the next rule. You should be aware that the LOG target: ☞ Logs all traffic that matches the iptables rule in which it is located.168. If you want to log only unwanted traffic.0/24 \ -j SNAT -o eth1 --to-source 97. you have to add a matching rule with a DROP target immediately after the LOG rule.168.158.qxd 2/25/05 10:06 AM Page 234 234 Linux Firewalls Using iptables Chapter 14 # (Connections originating from the home network servers) iptables -t nat -A POSTROUTING -s 192.1.26 # POSTROUTING statements for Many:1 NAT # (Connections originating from the entire home network) iptables -t nat -A POSTROUTING -s 192.1. therefore.253.16Harrison_ch14.RELATED -j ACCEPT # Allow forwarding for all 1:1 NAT connections originating on # the Internet that have already passed through the NEW forwarding # statements above iptables -A FORWARD -t filter -i eth0 -m state \ --state ESTABLISHED. If you . One of the best methods is to log all dropped packets to the /var/log/messages file.

102 LEN=220 TOS=0x00 PREC=0x00 TTL=54 ID=30485 PROTO=UDP SPT=53 DPT=32820 LEN=200 ☞ Firewall denies Windows NetBIOS traffic (UDP port 138) Feb 23 20:43:08 bigboy kernel: IN=wlan0 OUT= MAC=ff:ff:ff:ff:ff:ff:00:06:25:09:6a:b5:08:00 SRC=192. This example logs a summary of failed packets to the file /var/log/messages. if it is not. “Simple Network Troubleshooting.81.255 LEN=241 TOS=0x00 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=138 DPT=138 LEN=221 ☞ Firewall denies Network Time Protocol (NTP UDP port 123) Feb 23 20:58:48 bigboy kernel: IN= OUT=wlan0 SRC=192. OUTPUT.168.102 DST=207. FORWARD.168. the location on your network that could be causing the problem. you’ll find yourself logging both desired and unwanted traffic with no way of discerning between the two. #--------------------------------------------------------------# Log and drop all other packets to file /var/log/messages # Without this we could be crawling around in the dark #--------------------------------------------------------------iptables -A OUTPUT -j LOG iptables -A INPUT -j LOG iptables -A FORWARD -j LOG iptables -A OUTPUT -j DROP iptables -A INPUT -j DROP iptables -A FORWARD -j DROP Here are some examples of the output of this file: ☞ Firewall denies replies to DNS queries (UDP port 53) destined to server 192.93. and NAT related statements.100 DST=192.1.102 on the home network Feb 23 20:33:50 bigboy kernel: IN=wlan0 OUT= MAC=00:06:25:09:69:80:00:a0:c5:e1:3e:88:08:00 SRC=192.” to determine whether the data is reaching your firewall at all and.1.168. therefore. then you should focus on the INPUT and OUTPUT statements If nothing shows up in the logs.168.1. then follow the steps in Chapter 4.42.1.1. you should check your INPUT. You can use the contents of this file to determine which TCP/UDP ports you need to open to provide access to specific traffic that is currently stopped.16Harrison_ch14.168. If the firewall’s IP address is involved.qxd 2/25/05 10:06 AM Page 235 Troubleshooting iptables 235 don’t.200. because by default iptables doesn’t state why the packet was logged in its log message.113 LEN=76 TOS=0x10 PREC=0x00 TTL=64 ID=0 DF PROTO=UDP SPT=123 DPT=123 LEN=56 The traffic in all these examples isn’t destined for the firewall. .30 DST=192.

You can then use the logging output in /var/log/messages to make sure that the translations are occurring correctly and iptables isn’t dropping the packets after translation occurs. and. restricted and monitored physical access to your servers.d/iptables script running without the typical [OK] or [FAILED] messages. reliable power and cooling. regular application patching. then symptoms include the firewall status always being stopped and the /etc/init. secured cabling. Security should be viewed as anything that contributes to the desired risk-free functioning of your site.qxd 2/25/05 10:06 AM Page 236 236 Linux Firewalls Using iptables Chapter 14 As a general rule. Basic NAT testing requires you to ask a friend to try to connect to your home network from the Internet. you won’t be able to access the public NAT IP addresses from servers on your home network. running the service iptables save command before restarting won’t help either. probably most importantly. iptables Won’t Start The iptables startup script expects to find the /etc/sysconfig/iptables before it starts.16Harrison_ch14. Web site security involves not just protection from corrupted packets or maliciously overwhelming volumes of traffic. well trained and motivated employees. and it is well worth the money to invest in and learn from a book that specializes in the topic. If you have just installed iptables and have never applied a policy. then you will face this problem. but a firewall is never sufficient. enforced password policies. If none exists. . Unfortunately. You have to create this file: [root@bigboy tmp]# service iptables start [root@bigboy tmp]# [root@bigboy tmp]# touch /etc/sysconfig/iptables [root@bigboy tmp]# chmod 600 /etc/sysconfig/iptables [root@bigboy tmp]# service iptables start Applying iptables firewall rules: [ OK ] [root@bigboy tmp]# CONCLUSION A firewall is a critical part of any establishment that connects to an unprotected network such as the Internet. redundant hardware. but also involves daily data backups to help recovery from device failures.

most serveroriented operating systems usually include an FTP server application as part of the software suite. Most Web-based download sites use the built-in FTP capabilities of Web browsers. and. It operates using two connection channels: 237 .17Harrison_ch15. This chapter will show you how to convert your Linux box into an FTP server using the default Very Secure FTP Daemon (VSFTPD) package included in Fedora.qxd 2/25/05 10:06 AM Page 237 C H A P T E R 15 Linux FTP Server Setup IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ FTP Overview Problems with FTP and Firewalls How to Download and Install VSFTPD How to Get VSFTPD Started Testing the Status of VSFTPD The vsftpd. FTP OVERVIEW FTP relies on a pair of TCP ports to get the job done. Linux is no exception.conf File FTP Security Issues Troubleshooting FTP Tutorial Conclusion The File Transfer Protocol (FTP) is one of the most common means of copying files between servers over the Internet. therefore.

In active FTP. in which general access is provided to the FTP server using a well known universal login method. For passive FTP. in which files are transferred using the username and password of a regular user FTP server.1 Active and passive FTP. TCP Port 21: All commands you send. and anonymous FTP. go over the control connection. there are two additional types of FTP: regular FTP.17Harrison_ch15. the connection is initiated from the FTP client. Types of FTP From a networking perspective. there are several varieties of FTP. Take a closer look at each type. . as well as the FTP server’s responses to those commands. In addition to these channels. TCP Port 20: This port is used for all subsequent data transfers between the client and server.1. These are illustrated in Figure 15.qxd 2/25/05 2:40 PM Page 238 238 Linux FTP Server Setup Chapter 15 ☞ FTP control channel. From a user management perspective. ☞ FTP data channel. the two main types of FTP are active and passive. Active FTP FTP Control Connection to Port 21 from High Port High Port Client Computer FTP Data Connection Initiation from Port 20 on Server to High Port on Client Server Port 20 Passive FTP FTP Control Connection to Port 21 from High Port High Port Client Computer Server High Port FTP Data Connection Initiation from High Port on Client to High Port on the Server Figure 15. the FTP server initiates a data transfer connection back to the client. but any data sent back (such as ls directory lists or actual file data in either direction) will go over the data connection.

Passive FTP should be viewed as the server never making an active attempt to connect to the client for FTP data transfers. the client initiates the data transfer connections to the server. Your commands such as ls and get are sent over this connection. as it selects port 20 as its source port (not a random high port that’s greater than 1024) and connects back to the client on a random high port that has been pre-negotiated on the port 21 control connection. . Passive FTP Passive FTP works differently: 1. The source port of these data transfer connections is always port 20 on the server. Whenever the client requests data over the control connection. not the port 21 control connection. therefore. Your client connects to the FTP server by establishing an FTP control connection to port 21 of the server. Thus the ls listing that you asked for comes back over the port 20 to high port connection. the VSFTPD package allows regular Linux users to copy files to and from their home directories with an FTP client using their Linux usernames and passwords as their login credentials. As Windows defaults to active FTP and Linux defaults to passive. and the destination port is a high port (greater than 1024) on the client. FTP active mode. Your commands such as ls and get are sent over that connection. Whenever the client requests data over the control connection. Because the client always initiates the required connections. Regular FTP By default. passive FTP works better for clients protected by a firewall. 2. the server initiates data transfer connections back to the client. transfers data in a counter intuitive way to the TCP standard. Your client connects to the FTP server by establishing an FTP control connection to port 21 of the server. Active FTP may fail in cases where the client is protected from the Internet via many to one NAT (masquerading).qxd 2/25/05 2:40 PM Page 239 FTP Overview 239 Active FTP The sequence of events for active FTP is: 1. 2. The source port of these data transfer connections is always a high port on the client with a destination port of a high port on the server.17Harrison_ch15. 3. you’ll probably have to accommodate both forms when deciding upon a security policy for your FTP server. because the firewall will not know which of the many servers behind it should receive the return connection.

dir. Anonymous FTP allows you to avoid this difficulty. because firewalls are designed to limit data flows to predictable TCP ports and FTP uses a wide range of unpredictable TCP ports. enabling you to restrict the addition of new files to your system to authorized personnel.17Harrison_ch15. however. “Codes. Client Protected by a Firewall Problem Typically firewalls don’t allow any incoming connections at all. as you’ll see later. Unlike regular FTP where you login with a preconfigured Linux username and password. VSFTPD can be configured to support user-based and/or anonymous FTP in its configuration file. PROBLEMS WITH FTP AND FIREWALLS FTP frequently fails when the data has to pass through a firewall. “Installing RPM Software. Note Appendix II. and Configurations. With this type of FTP failure. Anonymous FTP Anonymous FTP is the choice of Web sites that need to exchange files with numerous unknown remote users. Scripts. The disadvantage of regular FTP is that it isn’t suitable for general download distribution of software as everyone either has to get a unique Linux user account or has to use a shared username and password. or get commands. The connection then appears to hang. You have a choice of methods to overcome this. The reason is that the firewall is blocking the return connection from the server to the client (from port . Once logged into a VSFTPD server. anonymous FTP requires only a username of anonymous and your e-mail address for the password. which frequently blocks active FTP from functioning.” using anonymous FTP as a remote user is fairly straightforward.qxd 2/25/05 10:06 AM Page 240 240 Linux FTP Server Setup Chapter 15 VSFTPD also has the option of allowing this type of access to only a group of Linux users. the active FTP connection appears to work when the client initiates an outbound connection to the server on port 21.” contains examples of how to configure the VSFTPD Linux firewall to function with both active and passive FTP. As seen in Chapter 6. you automatically have access to only the default anonymous FTP directory (/var/ftp in the case of VSFTPD) and all its subdirectories. as soon as you use the ls. Common uses include downloading software updates and MP3s and uploading diagnostic information for a technical support engineers’ attention.

qxd 2/25/05 10:06 AM Page 241 Problems with FTP and Firewalls 241 20 on the server to a high port on the client). High 20 High High New Established3 New Established3 2 In some cases. or network. the FTP connection from the client doesn’t appear to work at all for both active and passive FTP. not just a specific client.1 Method Client Protected by Firewall: Required Rules for FTP Destination Port 21 High Connection Type New Established3 Source Source Destination Address Port Address Allow outgoing control connections to server Control FTP client/ High1 FTP server2 channel network FTP server2 21 FTP client/ network Allow the client to establish data channels to remote server Active FTP server2 20 FTP client/ FTP network FTP client/ High FTP server2 network Passive FTP client/ High FTP server2 FTP network FTP server2 High FTP client/ network 1 Greater than 1024. If a firewall allows all outbound connections to the Internet. Solution Table 15. Table 15. 3 Many home-based firewall routers automatically allow traffic for already established connections.17Harrison_ch15. . This rule may not be necessary in all cases. When an incorrectly configured firewall protects an FTP server. Server Protected by a Firewall Problem Typically. firewalls don’t let any connections come in at all.2 outlines the general rules needed to allow FTP servers through a firewall. then passive FTP clients behind a firewall will usually work correctly as the clients initiate all the FTP connections. server.1 shows the general rules you’ll need to allow FTP clients through a firewall. you may want to allow all Internet users to have access. Solution Table 15.

2 Method Server Protected by Firewall: Required Rules for FTP Destination Port 21 High Connection Type New Established3 Source Source Destination Address Port Address Allow incoming control connections to server FTP server Control FTP client/ High2 1 channel network FTP server 21 FTP client/ network1 Allow server to establish data channel to remote client Active FTP server 20 FTP client/ FTP network1 FTP client/ High FTP server network1 Passive FTP client/ High FTP server FTP network1 FTP server High FTP client/ network1 High 20 High High New Established3 New Established3 1 In some cases. If you need a refresher. This rule may not be necessary in all cases. remember that the VSFTPD RPM’s filename usually starts with the word “vsftpd” followed by a version number. Chapter 6 covers how to do this in detail.17Harrison_ch15. 2 Greater than 1024. HOW TO DOWNLOAD AND INSTALL VSFTPD Most Red Hat and Fedora Linux software products are available in the RPM format.2.i386. or restart VSFTPD after booting using these commands: [root@bigboy tmp]# service vsftpd start [root@bigboy tmp]# service vsftpd stop [root@bigboy tmp]# service vsftpd restart . as in vsftpd-1.rpm. 3 Many home-based firewall routers automatically allow traffic for already established connections. When searching for the file. Downloading and installing RPMs isn’t hard. server. It is best to use the latest version of VSFTPD. you may want to allow all Internet users to have access. stop. or network.qxd 2/25/05 10:06 AM Page 242 242 Linux FTP Server Setup Chapter 15 Table 15. HOW TO GET VSFTPD STARTED You can start.1-5. not just a specific client.

” You can find a full description of how to configure these versions of Linux for VSFTPD in Appendix III. which lists all the TCP and UDP ports on which the server is listening for traffic. This file uses a number of default settings you need to know about: ☞ VSFTPD runs as an anonymous FTP server: Unless you want any remote user to log into to your default FTP directory using a username of anonymous and a password that’s the same as their e-mail address. so you’ll have to restart VSFTPD each time you edit the file in order for the changes to take effect. use the chkconfig command: [root@bigboy tmp]# chkconfig vsftpd on Note In Red Hat Linux version 8. ☞ VSFTPD allows only anonymous FTP downloads to remote users. which is covered in Chapter 16. “Fedora Version Differences.conf File 243 To configure VSFTPD to start at boot. and XINETD. not uploads from them: You can change this by modifying the anon_upload_enable directive shown later.conf configuration file only when it starts.CONF FILE VSFTPD reads the contents of its vsftpd.” TESTING THE STATUS OF VSFTPD You can always test whether the VSFTPD process is running by using the netstat -a command. . I suggest turning this off. “TELNET. You’ll also need to simultaneously enable local users to be able to log in by removing the comment symbol (#) before the local_enable instruction. This example shows the expected output: [root@bigboy root]# netstat -a | grep ftp tcp 0 0 *:ftp [root@bigboy root]# *:* LISTEN If VSFTPD wasn’t running. VSFTPD operation is controlled by the xinetd process. THE VSFTPD. there would be no output at all. TFTP.0 and earlier.17Harrison_ch15.qxd 2/25/05 10:06 AM Page 243 The vsftpd. You can set the configuration file’s anonymous_enable directive to no to disable this feature.

log # The directory which vsftpd will try to change # into after an anonymous login. #anon_mkdir_write_enable=YES # Activate logging of uploads/downloads. #anon_upload_enable=YES # Uncomment this if you want the anonymous FTP user to be able to create # new directories.qxd 2/25/05 10:06 AM Page 244 244 Linux FTP Server Setup Chapter 15 ☞ VSFTPD doesn’t allow anonymous users to create directories on your FTP server: You can change this by modifying the anon_mkdir_write_enable directive. xferlog_enable=YES # You may override where the log file goes if you like. . you will # obviously need to create a directory writable by the FTP user. # (Needed even if you want local users to be able to upload files) write_enable=YES # Uncomment to allow the anonymous FTP user to upload files. local_enable=YES # Uncomment this to enable any form of FTP write command. remove or add the # at the beginning of the appropriate line. There is always the risk with anonymous FTP that users will discover a way to write files to your anonymous FTP directory. #xferlog_file=/var/log/vsftpd.log log file: You can change this by modifying the xferlog_file directive. Also. You run the risk of filling up your /var partition if you use the default setting. The configuration file is fairly straightforward as you can see in the snippet: # Allow anonymous FTP? anonymous_enable=YES # Uncomment this to allow local users to log in. It is best to make the anonymous FTP directory reside in its own dedicated partition. ☞ VSFTPD expects files for anonymous FTP to be placed in the /var/ftp directory: You can change this by modifying the anon_root directive. ☞ VSFTPD logs FTP access to the /var/log/vsftpd. (Default = /var/ftp) #anon_root=/data/directory To activate or deactivate a feature. # The default is shown# below.17Harrison_ch15. This only # has an effect if global write enable is activated.

conf man FTP SECURITY ISSUES FTP has a number of security drawbacks. but you can overcome them in some cases.conf Options There are many other options you can add to this file: ☞ Limiting the maximum number of client connections (max_clients) ☞ Limiting the number of connections by source IP address (max_per_ip) ☞ Setting the maximum rate of data transfer per anonymous login (anon_max_rate) ☞ Setting the maximum rate of data transfer per non-anonymous login (local_max_rate) Descriptions on this and more can be found in the pages. The commands you need are: [root@bigboy tmp]# mkdir /var/ftp/pub/upload [root@bigboy tmp]# chmod 722 /var/ftp/pub/upload . and you can change the configuration to not display the FTP server’s software version information. The /etc/vsftpd. then you should create a write-only directory within /var/ftp/pub. though very convenient. You can restrict an individual Linux user’s access to non-anonymous FTP. Anonymous Upload If you want remote users to write data to your FTP server. vsftpd. As FTP doesn’t encrypt passwords. FTP logins and data transfers are not encrypted. but unfortunately.ftpusers file.qxd 2/25/05 10:06 AM Page 245 FTP Security Issues 245 Other vsftpd. it is a good idea to let these entries remain and add new entries for additional security.ftpusers File For added security. you may restrict FTP access to certain users by adding them to the list of users in the /etc/vsftpd.17Harrison_ch15. This will allow your users to upload but not access other files uploaded by other users. The VSFTPD package creates this file with a number of entries for privileged users that normally shouldn’t have FTP access. thereby increasing the risk of data or passwords being compromised.

. one of which is allowing numerous unknown users to download files. images. because you run the risk of accidentally allowing unknown persons to upload files to your server.17Harrison_ch15. This could make your user account vulnerable to an unauthorized attack from a person eavesdropping on the network connection. Connection problems could also be the result of typical network issues outlined in Chapter 4. which in turn can clog your server’s Internet access and drive up your bandwidth charges. TROUBLESHOOTING FTP You should always test your FTP installation by attempting to use an FTP client to log into your FTP server to transfer sample files. “Simple Network Troubleshooting. SCP does not support anonymous services. Follow the firewall rule guidelines to help overcome this problem.qxd 2/25/05 10:06 AM Page 246 246 Linux FTP Server Setup Chapter 15 FTP Greeting Banner Change the default greeting banner in the vsftpd. The most common sources of day-to-day failures are incorrect usernames and passwords. Secure Copy (SCP) and Secure FTP (SFTP) provide encryption and could be considered as an alternative to FTP for trusted users. This sort of unintended activity can quickly fill up your hard drive with illegal software. Initial setup failures could be caused by firewalls along the path between the client and server blocking some or all types of FTP traffic. however.” TUTORIAL FTP has many uses. a feature that FTP does support. and music for the world to download. You have to be careful.conf file to make it harder for malicious users to determine the type of system you have. The directive in this file is: ftpd_banner= New Banner Here Using SCP as Secure Alternative to FTP One of the disadvantages of FTP is that it does not encrypt your username and password. Typical symptoms of this are either connection timeouts or the ability to use the ls command to view the contents of a directory without the ability to either upload or download files.

Copy files to be downloaded by your users into the /home/ftp-docs directory. use /home/ftpusers and a user group name of ftp-users for the remote users: [root@bigboy tmp]# groupadd ftp-users [root@bigboy tmp]# mkdir /home/ftp-docs 5. and make their default directory /home/ftp-docs: [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user1 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user2 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user3 [root@bigboy tmp]# useradd -g ftp-users -d /home/ftp-docs user4 [root@bigboy tmp]# passwd user1 [root@bigboy tmp]# passwd user2 [root@bigboy tmp]# passwd user3 [root@bigboy tmp]# passwd user4 7. Change the permissions of the files in the /home/ftp-docs directory to read-only access by the group: [root@bigboy tmp]# chown root:ftp-users /home/ftp-docs/* [root@bigboy tmp]# chmod 740 /home/ftp-docs/* . anonymous FTP is not desired. In this case. local_enable=YES local_enable line 3. Create a user group and shared directory.qxd 2/25/05 10:06 AM Page 247 Tutorial 247 FTP Users with Read-Only Access to a Shared Directory In this example. [root@bigboy tmp]# service vsftpd start 4. Make the directory accessible to the ftp-users group: [root@bigboy tmp]# chmod 750 /home/ftp-docs [root@bigboy tmp]# chown root:ftp-users /home/ftp-docs 6. Disable anonymous FTP. Here are the steps: 1.17Harrison_ch15. but a group of trusted users need to have read-only access to a directory for downloading files.conf file: # Uncomment this to allow local users to log in. Enable individual logins by making sure you have the uncommented in the vsftpd. Add users.conf file: # Allow anonymous FTP? # anonymous_enable=YES anonymous_enable line in the 2. Comment out the vsftpd. 8. Start VSFTP.

ftp> As expected.0-1. Sample Login Session to Test Functionality Here is a simple test procedure you can use to make sure everything is working correctly: 1.conf file to no: write_enable = NO Remember.1.168.17Harrison_ch15. you can’t do an upload transfer of testfile to bigboy: ftp> put testfile local: testfile remote: testfile 227 Entering Passive Mode (192.100 (192.1.181. Remote system type is UNIX. Check for the presence of a test file on the FTP client server.1.100.168.rpm .1.173) 150 Here comes the directory listing.168.1.168. Password: 230 Login successful.100 Connected to 192.1.168. then you should set the write_enable line in your vsftpd. break me) Name (192. ftp> But we can view and download a copy of the VSFTPD RPM on the FTP server bigboy: ftp> ls 227 Entering Passive Mode (192.100:root): user1 331 Please specify the password.i386.168.0: beat me. Using binary mode to transfer files.1. -rwxr----.qxd 2/25/05 10:06 AM Page 248 248 Linux FTP Server Setup Chapter 15 Users should now be able to log in via FTP to the server using their new usernames and passwords.35. Connect to Bigboy via FTP: [root@smallfry tmp]# ftp 192.1 root root 0 Jan 4 09:08 testfile [root@smallfry tmp]# 2.100) 220 ready. [root@smallfry tmp]# ll total 1 -rw-r--r-.1 0 502 76288 Jan 04 17:06 vsftpd-1.100. you must restart VSFTPD for the configuration file changes to take effect.210) 553 Could not create file. Have fun.1. dude (vsFTPd 1. If you absolutely don’t want any FTP users to be able to write to any directory.

Although insecure.i386.100 (192.1.1.1.qxd 2/25/05 10:06 AM Page 249 Conclusion 249 226 Directory send OK. an anonymous FTP fails: [root@smallfry tmp]# ftp 192.168.1.1.168.0-1.1. ftp> quit 221 Goodbye. “Secure Remote Logins and File Copying.0-1.168.01. Password: 530 Login incorrect.168.1. you can make this a regular part of your FTP server’s operation.i386.499 secs (1.156) 150 Opening BINARY mode data connection for vsftpd-1. dude (vsFTPd 1.100.44.100:root): anonymous 331 Please specify the password.rpm. Login failed.rpm. You can find more information on it in Chapter 17.rpm vsftpd-1.168.1. [root@smallfry tmp]# Now that testing is complete.100 Connected to 192.0: beat me. then you should probably consider SCP as a possible alternative.1.rpm (76288 bytes). [root@smallfry tmp]# As expected.1. break me) Name (192. 226 File send OK. CONCLUSION FTP is a very useful software application that can have enormous benefit to a Web site or to collaborative computing in which files need to be shared between business partners.tmp remote: vsftpd-1.i386.01. If data encryption security is of great importance to you.rpm 227 Entering Passive Mode (192.100) 220 ready.5e+02 Kbytes/sec) ftp> exit 221 Goodbye. because FTP clients are a part of all operating systems and Web browsers. ftp> get vsftpd-1.” .tmp local: vsftpd-1. it is universally accessible.i386. 76288 bytes received in 0.1.i386.17Harrison_ch15.0-1.

17Harrison_ch15.qxd 2/25/05 10:06 AM Page 250 .

and restart xinetd after booting by using: [root@bigboy tmp]# service xinetd start [root@bigboy tmp]# service xinetd stop [root@bigboy tmp]# service xinetd restart xinetd RPM /etc/xinetd. TFTP. Fortunately you usually don’t have to edit this file.conf as 251 . and chkconfig controls its behavior at boot time. MANAGING The XINETD PROGRAMS is installed by default in Fedora Linux and uses its main configuration file. Controlling xinetd The scripts in the /etc/init.qxd 2/25/05 10:07 AM Page 251 C H A P T E R 16 TELNET.d directory control the starting and stopping of the xinetd daemon. instead they often offload a lot of this work to a program suite made just for this purpose: xinetd.18Harrison_ch16. so day-to-day xinetd operation is frequently limited to only starting and stopping xinetd-managed applications. and xinetd IN THIS CHAPTER ☞ Managing xinetd Programs ☞ TELNET ☞ Conclusion Many network-enabled Linux applications don’t rely on themselves to provide restricted access or bind to a particular TCP port. You can start. stop.

The TELNET server RPM is installed and disabled by default on Fedora Linux. xinetd configured to start at boot. TFTP. . the older TELNET application remains popular. One of the disadvantages of Telnet is that the data is sent as clear text. Each configuration file has a disable statement that you can set to yes or no. and xinetd Chapter 16 To get command. Using The TELNET Client The command to do remote logins via TELNET from the command line is simple. You don’t have to edit these files to activate or deactivate the application.d directory. Many network devices don’t have SSH clients.qxd 2/25/05 10:07 AM Page 252 252 TELNET. This means that it is possible for someone to use a network analyzer to peek into your data packets and see your username and password. which uses varying degrees of encryption. I’ll show you how to limit your exposure to TELNET’s insecurities later in this chapter. you can use the chkconfig [root@bigboy tmp]# chkconfig xinetd on Controlling xinetd-Managed Applications Xinetd-managed applications all store their configuration files in the /etc/xinetd. A more secure method for remote logins would be via Secure Shell (SSH). This governs whether xinetd is allowed to start them or not. In spite of this. You enter the word “telnet” and then the IP address or server name to which you want to connect.18Harrison_ch16. The chkconfig command does that for you automatically and also stops or starts the application accordingly too! Here is an example of the activation and deactivation of the Samba SWAT Web GUI management application: [root@bigboy tmp]# chkconfig swat on [root@bigboy tmp]# chkconfig swat off TELNET TELNET is a program that allows users to log into your server and get a command prompt just as if they were logged into the VGA console. making Telnet the only means of accessing other devices and servers from them.

[peter@smallfry peter]$ [peter@smallfry peter]$ netstat –nr Kernel IP routing table Destination Gateway Genmask Flags Iface 255. [root@bigboy tmp]# telnet 192. If you need a refresher.0..0.0.0 255.255.1.0 0. remember that the TELNET server RPM’s filename usually starts with telnet-server followed by a version number.105. as in telnet-server-0.0.rpm.18-14 (smallfry. [root@bigboy tmp]# MSS Window irtt 40 40 40 40 0 0 0 0 0 0 0 0 Installing the TELNET Server Software Older versions of Red Hat had the TELNET server installed by default.168.255 UH wlan0 192. .255.0 255. 05 January 2003) Login: peter Password: Last login: Fri Nov 22 23:29:44 on ttyS0 You have new mail.105 Trying 192. Downloading and installing RPMs isn’t hard. Linux 2.168.0.1. Connected to 192.1.168. Chapter 6.255.18Harrison_ch16. The user looks at the routing table and then logs out. “Installing RPM Software.255.qxd 2/25/05 2:41 PM Page 253 TELNET 253 Here is an example of someone logging into a remote server named smallfry from server Bigboy. Fedora Linux doesn’t do this.0.255.17-28.0.i386.168.0 0. and you will have to install it yourself.0.1.105. Most Red Hat and Fedora Linux software products are available in the RPM format..” covers how to do this in detail.0 255.0.0 U wlan0 127.0 U lo 0.0 UG wlan0 [peter@smallfry peter]$ exit logout Connection closed by foreign host.1 0.0.0.0.org) (10:35 on Sunday. Escape character is ‘^]’.my-web-site.0.255.4.1.0.168. When searching for the file.0 192.255 0.

then there will be no response. but it makes it less likely to be detected as TELNET traffic. [root@bigboy tmp]# netstat -a | grep telnet tcp 0 0 *:telnet LISTEN [root@bigboy tmp]# *:* You can also use the chkconfig will be started on the next reboot: --list command to verify that TELNET [root@bigboy tmp]# chkconfig --list | grep telnet telnet: on [root@bigboy tmp]# Stopping a TELNET Server Use the chkconfig command to deactivate TELNET. you should also try to ensure that TELNET sessions run over secure internal networks or across VPNs to reduce the risk of exposing sensitive data to unauthorized eyes. Remember that this isn’t a foolproof strategy. use the chkconfig command to activate TELNET: [root@bigboy tmp]# chkconfig telnet on You can test whether the TELNET process is running with the netstat command. Call it # Local services stelnet 7777/tcp telnet # “secure” . good port scanning programs can detect TELNET and other applications running on alternative ports.18Harrison_ch16. Let Telnet Listen on Another TCP Port Letting TELNET run on an alternate TCP port doesn’t encrypt the traffic. Check out some other options.qxd 2/25/05 10:07 AM Page 254 254 TELNET. The steps are: 1. and xinetd Chapter 16 Setting Up a TELNET Server To set up a TELNET server. even after the next reboot: [root@bigboy tmp]# chkconfig telnet off Basic TELNET Security There are a number of things you can do to improve the security of TELNET. For example. which is used to check the TCP/UDP ports on which your server is listening. If it isn’t running. Edit your stelnet. /etc/services file. TFTP. and add an entry for a new service.

Here’s how: 1.100.0:7777 0.d/stelnet: /etc/xinetd.0:* [root@bigboy tmp]# LISTEN You should now be able to log into the new stelnet server on port 7777.d/telnet file separated by spaces: # default: on # description: The telnet server serves telnet sessions .168.0.d/stelnet file. Use chkconfig to activate stelnet: [root@bigboy tmp]# chkconfig stelnet on 5.telnetd log_on_failure += USERID disable = no port = 7777 } 4.1. service stelnet { flags = REUSE socket_type = stream wait = no user = root server = /usr/sbin/in.0.521 on an i686 login: Let TELNET Allow Connections from Trusted Addresses You can restrict TELNET logins’ access to individual remote servers by using the only_from keyword in the TELNET configuration file.8-1. Connected to 192. This is done using the telnet command with the TCP port as the second argument. Copy the TELNET configuration file called call it /etc/xinetd.100 7777 Trying 192. Make the new service stelnet. Fedora Core release 2 (Tettnang) Kernel 2.d/telnet /etc/xinetd... Check to make sure your server is now listening on port 7777 with the netstat command: [root@bigboy tmp]# netstat -an | grep 777 tcp 0 0 0.0.0.6. Add a list of trusted servers to the /etc/xinetd. and add a port statement for TCP port 7777: # default: on # description: The telnet server serves telnet sessions # unencrypted username/password pairs for authentication.1.1. Edit the new /etc/xinetd. Escape character is ‘^]’.168. [root@smallfry tmp]# telnet 192.100.168.qxd 2/25/05 10:07 AM Page 255 TELNET 255 2.d/telnet and [root@bigboy tmp]# cp /etc/xinetd.d/stelnet 3.18Harrison_ch16.

telnetd log_on_failure += USERID disable = no only_from = 192. Test the TELNET session.1.100 Trying 192. such as files containing updated access control lists (ACLs) that restrict access to networks and even the regular application of new passwords.0. Finally. which is why the topic is covered here.200 } 2.. Connected to 192. Escape character is ‘^]’. Connection closed by foreign host. TFTP can be used with great versatility as a network management tool and not just for saving files.100.168.168.1. a leading networking hardware manufacturer.18Harrison_ch16. Restart TELNET: [root@bigboy tmp]# chkconfig telnet off [root@bigboy tmp]# chkconfig telnet on 3. TFTP.1. [root@smallfry tmp]# TFTP Many networking equipment manufacturers allow you to back up live configurations of their devices to centralized servers via the TFTP protocol.100 127.” [root@smallfry tmp]# telnet 192. .168. and xinetd Chapter 16 # unencrypted username/password pairs for authentication.0. they can be used to upload even partial configurations.1.1 192.168. TFTP servers can be used to upload new configurations to replacement devices after serious hardware failures. The provided TFTP examples use equipment from Cisco Systems. Servers that are not on the trusted list get the message “Connection closed by foreign host.100.168.1. service telnet { flags = REUSE socket_type = stream wait = no user = root server = /usr/sbin/in.qxd 2/25/05 10:07 AM Page 256 256 TELNET. They also can be used for uploading new versions of software to be run as network devices.. TFTP may not be an application used regularly in a home. but it will become increasingly important in an expanding small office/home office (SOHO) environment.

33-3. TFTP files of increasing size could eventually fill the partition. and then creates a symbolic link that makes this directory appear to also be the /tftpboot directory: [root@bigboy tmp]# mv /tftpboot /var [root@bigboy tmp]# ln -s /var/tftpboot /tftpboot You must restart xinetd for the new configuration to take effect: [root@bigboy tmp]# chkconfig tftp on Each device must have a configuration file in the /tftpboot directory. Change this in the /etc/xinetd. Here’s an example of what to do for a SOHO firewall named pixfw and a configuration filename that matches Cisco’s standard naming scheme of devicename-config: [root@bigboy [root@bigboy [root@bigboy total 1631 -rw-rw-rw. Configuring the TFTP Server The procedure to set up a TFTP Server is straightforward. as in: tftp-server-0. Downloading and installing RPMs isn’t hard.i386. the TFTP application expects files to be located in the /tftpboot directory. If you need a refresher.rpm. By default. or create your own directory just for this purpose and create a /tftpboot symbolic link to it.1 [root@bigboy tmp]# touch /tftpboot/pixfw-config tmp]# chmod 666 /tftpboot/pixfw-config tmp]# ll /tftpboot/ root root 3011 Oct 29 14:09 pixfw-config tmp]# . remember that the TFTP server RPM’s filename usually starts with word tftp-server and is followed by a version number.d/tftp file via the server_args option. When searching for the file.18Harrison_ch16. This example creates a new tftpboot directory in the /var partition. affecting your ability to install new software or even the overall performance of your system. It is usually best to place the TFTP files in a partition other than the root partition.qxd 2/25/05 2:42 PM Page 257 TFTP 257 Installing the TFTP Server Software Most Red Hat and Fedora Linux software products are available in the RPM format. see Chapter 6.

The examples assume that the TFTP server’s IP address is 192.168.. TFTP write ‘/pixfw-config’ at 192.1.. Save the configuration to non-volatile memory: pixfw# write memory Building configuration. and xinetd Chapter 16 You can test whether the TFTP process is running with the netstat command.168. Save the configuration to the TFTP server: pixfw# write network Building configuration.168. Cryptochecksum: 3af43873 d35d6f06 51f8c999 180c2342 [OK] pixfw# 3. get into Enable mode. and then enter the TFTP commands to initially configure TFTP: pixfw> enable Password: ******** pixfw# configure terminal pixfw(config)# tftp-server inside 192. TFTP. Cisco PIX Firewall Follow these steps on a PIX firewall: 1..100 on interface 1 [OK] pixfw# Your firewall configuration has now been successfully saved for later use in the event of unrecoverable human error or hardware failure. and then enter the write net TFTP command: . which is used to check the TCP/UDP ports on which your server is listening.. Log onto the device. you need to log onto the device.100 /pixfw-config pixfw(config)# exit 2.1.100. Cisco Switch Running CATOS To save the configuration of a Catalyst-series switch running CATOS.qxd 2/25/05 10:07 AM Page 258 258 TELNET.1. [root@bigboy tmp]# netstat -a | grep tftp udp 0 0 *:tftp [root@bigboy tmp]# *:* Saving Cisco Configurations to the TFTP Server You’ll now have to configure your Cisco router/firewall to use the TFTP server. If it isn’t running then there will be no response.18Harrison_ch16. get into Enable mode.

(\) 100% Connecting (/) Completed successfully. (30907 bytes) ciscoswitch> (enable) Cisco Router To save the configuration of a router.168.168. ciscocss# exit Cisco Local Director To save the configuration of a Cisco Local Director load balancer.100] Name of configuration file?[ciscoswitch-config] Upload configuration to ciscoswitch-config on 192.1.168. and then enter the TFTP commands: ciscorouter> enable ciscorouter# write net Remote host [192..1.100? [confirm] y ciscorouter# exit Cisco CSS 11000 Arrowpoints To save the configuration of a Cisco CSS-series load balancer. log onto the device. Finished network upload..100 (y/n) [n]? y . switch to Configure mode.168. writing configuration to //ciscold-config on 192. IP address or name of remote host? [192.. [OK] ciscold# exit .100 ciscocss-config Working..100]? 192. Use ‘write network all’ to show both default and non-default configurations.100 Name of configuration file to write [ciscorouter-config]? ciscorouterconfig Write file ciscorouter-config on host 192. log onto the device.qxd 2/25/05 10:07 AM Page 259 TFTP 259 ciscoswitch> (enable) wr net This command shows non-default configurations only.1.1. get into Enable mode. switch to Configure mode.18Harrison_ch16.1....168..1.. and then enter the TFTP commands: ciscold> ena Password: ciscold# write net 192.1.1.100:69 . get into Enable mode..168.. log onto the device and enter the TFTP commands as seen below: ciscocss# copy running-config tftp 192.168..168.100 ciscold-config Building configuration..

In this example.1.1.0 0.0.26521/52224 bytes] ciscorouter# Here’s a sample copy tftp: running-config command: .0.168.18Harrison_ch16. and it looks like this: ! ! Set the console password ! line con 0 password 7 $1$qDwqJEjunK$tuff0HE/g31/b7G/IZ ! ! Delete and recreate access list #10 ! no access-list 10 access-list 10 permit 192.100/config. So if the file was located in the /tftpboot directory.100 (via FastEthernet0/0): !!!!!! [OK . but if it were in the /tftboot/configs directory. a small file containing a new encrypted password and access control list is uploaded from the TFTP server and inserted into a router configuration.168.10.168.255 Procedure to Upload Configuration File Uploading the file can be done using either the copy tftp: running-config or the older configure network commands. it would be referred to as config. TFTP.file Configure using tftp://192.168.0 0.1.file from 192.0.file.file.0.1. you are prompted for the IP address of the TFTP server and the name of the file with the configuration commands.qxd 2/25/05 10:07 AM Page 260 260 TELNET. Consider this sample configure network command: ciscorouter>ena Password: ciscorouter#configure network Host or network configuration file [host]? This command has been replaced by the command: ‘copy <url> system:/running-config’ Address or name of remote host []? 192.file. and xinetd Chapter 16 Uploading Cisco Configurations from the TFTP Server From time to time you may have to upload configurations from your TFTP server to your network equipment.168. Sample Upload Configuration File The configuration file is named config.100 Source filename []? config.255 access-list 10 permit 192.file? [confirm] Loading config. The filename provided is always relative to the /tftpboot directory. it would be referred to as /tftboot/configs/config. In both cases.

This includes IP addresses. If the replacement unit is identical. ! Enter the commands to provide a bare minimum of connectivity to ! your TFTP server here. you can use TFTP to completely restore the configuration to the replacement device. The procedure for restoring your configuration is simple: 1.. 4. Give your router the bare minimum configuration that allows it to ping your TFTP server (no access controls or routing protocols). 3.1.168. 6.file from 192... where you have to replace a router completely.. Use the copy command to copy the backup configuration from the TFTP server to your startup configuration in NVRAM. Using TFTP to Restore Your Router Configuration In disastrous cases.. Log into the router via the console and verify the configuration is OK. 7. .912 secs (26521 bytes/sec) ciscorouter# Always remember to permanently save your configurations to nonvolatile RAM (NVRAM) afterwards with the write memory or copy running-config startup-config. a default route ! and the TFTP setup commands.. 5. Disconnect the router from the network. On rebooting. the router will copy the startup configuration stored in NVRAM into a clean running configuration environment. Loading config.100 (via FastEthernet0/0): !!!!!! [OK . Reconnect the router to the networks on which it was originally connected.26521/52224 bytes] 26521 bytes copied in 1.168.100 Source filename []? config. but expect to edit it if the interface names and software versions are different.qxd 2/25/05 10:07 AM Page 261 TFTP 261 ciscorouter#copy tftp: running-config Address or name of remote host []? 192.18Harrison_ch16. The commands you need are: ciscorouter> enable Password: ******** ciscorouter# write erase .168.. 2. Connect your router to the local network of the TFTP server.file Destination filename [running-config]? Accessing tftp://192.file. .1.1. Reload the router without saving the live running configuration to overwrite the startup configuration.100/config.. then you need to do very little editing of the saved configuration file. .

TFTP. and valuable usernames and passwords are exchanged. making eavesdropping easier and more lucrative for the hacker. TFTP sessions don’t even need a username and password. and xinetd Chapter 16 ..18Harrison_ch16.qxd 2/25/05 10:07 AM Page 262 262 TELNET.. such as expect. TELNET is a greater security risk as the connections are longer. They both have the shortcoming of not encrypting their data and therefore need to be used on secured networks for improved security.” . always use it with great care. I’d suggest that you use an encrypted TELNET replacement whenever possible. “Secure Remote Logins and File Copying. One such product. will be covered in Chapter 17. SSH. and the TFTP server process overwrites any existing file beneath its root directory in keeping with the instructions of the network engineer. ciscorouter# copy tftp:file-name startup-config ciscorouter# reload Please be aware that the write erase command erases your NVRAM startup configuration. CONCLUSION TELNET and TFTP are important applications in an overall systems administration strategy. Making mistakes with TFTP can be fairly easy to do. and you may want to consider automating the process by using a helper application.

qxd 2/25/05 10:08 AM Page 263 C H A P T E R 17 Secure Remote Logins and File Copying IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ A Quick Introduction to SSH Encryption Starting OpenSSH Testing the Status of SSH The /etc/ssh/sshd_config File Using SSH to Log Into a Remote Machine What to Expect with Your First Login Deactivating TELNET After Installing SSH Executing Remote Commands on Demand with SSH SCP: A Secure Alternative to FTP SFTP: Another Secure Alternative to FTP Using SSH and SCP Without a Password Conclusion One of the most popular file transfer and remote login Linux applications is OpenSSH.19Harrison_ch17. and Secure Shell (SSH) is often used as a stealthy alternative to TELNET. OpenSSH isn’t limited to Linux. The OpenSSH Secure Copy (SCP) and Secure FTP (SFTP) programs are secure replacements for FTP. SSH and SCP clients are available for most operating systems including Windows. which provides a number of ways to create encrypted remote terminal and file transfer connections between clients and servers. A QUICK INTRODUCTION TO SSH ENCRYPTION Data encryption is accomplished by using special mathematical equations to scramble the bits in a data stream to make it unreadable to anyone who does 263 .

and authorized_keys stores all the authorized public keys from remote hosts that may log into your account without the need for passwords (more on this later).qxd 2/25/05 10:08 AM Page 264 264 Secure Remote Logins and File Copying Chapter 17 not have access to the corresponding decryption equation. In this case. reinstalling the operating system or upgrading the SSH application might regenerate the keys. Here’s how it’s done with SSH.ssh/known_hosts along with the IP address that provided it. Your server’s own public and private SSH keys are stored in the /etc/ssh/ directory. This creates a situation in which the computers at each end of the SSH connection have each other’s public keys and are able to decrypt the data sent from the other end of the encrypted link or “tunnel”. the SSH connection is established. SSH uses the concept of randomly generated private and public keys to do its encryption. key changes can be caused by someone trying some sort of cyber attack as well. . The ~/ notation is a universally accepted way of referring to your home directory and is recognized by all Linux commands. You can recover the original data only if you have access to this key and the corresponding programs. For example. Of course. These files are also stored in your ~/.19Harrison_ch17. as are all files and directories whose names begin with a period. you are prompted as to whether you want to accept the download of the server’s public key before you can proceed. and then the client automatically sends its public key which the server uses to match against a predefined list in the user’s directory. The keys are usually created only once. The id_dsa and id_dsa. A successful exchange of encrypted data requires the receiver to have a copy of the sender’s public key beforehand.ssh directory and need to be specially generated. Note The . Always investigate changes to be safe. When you log into an SSH server. Linux also uses other key files at the user level to provide the capability of password-less logins and file copying to remote servers using SSH and SCP.pub files are your private and public keys respectively.ssh directory is a hidden directory. The process is usually made even harder through the use of an encryption key that is used to modify the way the equations do the scrambling. Data encryption helps to prevent unauthorized users from having access to the data. The ls -a command lists all normal and hidden files in a directory. then SSH knows that something is wrong. The SSH client’s public key is uploaded to the server at the same time. If there is a match then the login is authorized. All the public keys that an SSH client’s Linux user encounters are stored in a file named ~/. but you have the option of regenerating them should they become compromised. If a key and IP address no longer match.

0. [root@bigboy tmp]# chkconfig sshd on sshd You can also start.qxd 2/25/05 10:08 AM Page 265 The /etc/ssh/sshd_config File 265 STARTING OPENSSH OpenSSH is installed by default during Linux installations. Take a look at a snippet from configuration: # The strategy used for options in the default sshd_config shipped with # OpenSSH is to specify options with their default value where # possible. #Port 22 #Protocol 2. Uncommented options change a # default value. Because SSH and SCP are part of the same application.0.19Harrison_ch17. By default SSH listens on all your NICs and uses TCP port 22. [root@bigboy tmp]# service sshd start [root@bigboy tmp]# service sshd stop [root@bigboy tmp]# service sshd restart Remember to restart the SSH process every time you make a change to the configuration files for the changes to take effect on the running process.1 #ListenAddress 0. You can configure SSH to start at boot by using the chkconfig command. THE /ETC/SSH/SSHD_CONFIG FILE The SSH configuration file is called /etc/ssh/sshd_config. but leave them commented. stop. they share the same configuration file and are governed by the same /etc/init. and restart SSH after booting by running the initialization script.d/sshd startup script.0 #ListenAddress :: . TESTING THE STATUS OF SSH You can test whether the SSH process is running with: [root@bigboy tmp]# pgrep sshd You should get a response of plain old process ID numbers.

Restart SSH: [root@bigboy tmp]# service sshd restart 4.qxd 2/25/05 10:08 AM Page 266 266 Secure Remote Logins and File Copying Chapter 17 SSH Versions 1 and 2 The original encryption scheme of SSH was adequate for its time but was eventually found to have a number of limitations.1.0:* [root@bigboy root]# LISTEN Next. Remember to restart SSH to make this take effect.100:435 0. using grep to filter out everything that doesn’t have the string “435”: [root@bigboy root]# netstat -an | grep 435 [root@bigboy root]# 2.0. Always force your systems to operate exclusively with version 2 by setting the protocol statement in the /etc/ssh/sshd_config file to 2. because good network scanning programs can detect SSH running on alternative ports. This is a rudimentary precaution only. such as port 435. . you’ll discover how to actually log into systems using SSH.168. No response allows us to proceed. pick another port and try again: Port 435 3. What you need to do is: 1.0. The answer to these was version 2. then you can change port 22 to a location that won’t interfere with other applications on your system. Use the netstat command to make sure your system isn’t listening on port 435. Change the Port line in /etc/ssh/sshd_config to mention 435 and remove the # at the beginning of the line.19Harrison_ch17. Check to ensure SSH is running on the new port: [root@bigboy root]# netstat -an | grep 435 tcp 0 0 192. If port 435 is being used. # # File: /etc/ssh/sshd_config # Protocol 2 Change the TCP Port on Which SSH Listens If you are afraid of people trying to hack in on a well known TCP port.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 267

What to Expect with Your First Login

267

USING SSH TO LOG INTO A REMOTE MACHINE
Using SSH is similar to TELNET. To log in from another Linux box use the ssh command with a -l to specify the username you wish to log in as. If you leave out the -l, your username will not change. Here are some examples for a server named Smallfry in your /etc/hosts file. You can also use the username@remote_server format as an alternative. If you are user root and you want to log into Smallfry as yourself, use the command:
[root@bigboy tmp]# ssh smallfry

User root can also log into Smallfry as user peter via the default port 22:
[root@bigboy tmp]# ssh -l peter smallfry

or via port 435 using the username@remote_server alternative login format:
[root@bigboy tmp]# ssh -p 435 peter@smallfry

WHAT TO EXPECT WITH YOUR FIRST LOGIN
The first time you log in, you get a warning message saying that the remote host doesn’t know about your machine and prompting you to store a copy of the remote host’s SSH identification keys on your local machine. It will look something like this:
[root@bigboy tmp]# ssh smallfry The authenticity of host ‘smallfry (smallfry)’ can’t be established. RSA key fingerprint is 5d:d2:f5:21:fa:07:64:0d:63:1b:3b:ee:a6:58:58:bb. Are you sure you want to continue connecting (yes/no)? yes Warning: Permanently added ‘smallfry’ (RSA) to the list of known hosts. root@smallfry’s password: Last login: Thu Nov 14 10:18:45 2002 from 192.168.1.98 No mail. [root@smallfry tmp]#

The key is stored in your ˜/.ssh/known_hosts file and you should never be prompted for this again.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 268

268

Secure Remote Logins and File Copying

Chapter 17

SSH Failures Due to Linux Reinstallations If Linux or SSH is reinstalled on the remote server, then the keys are regenerated and your SSH client will detect that this new key doesn’t match the saved value in the known_hosts file. The SSH client will fail, erring on the side of caution to alert you to the possibility of a form of hacking attack:
[root@bigboy tmp]# ssh 192.168.1.102 @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ @ WARNING: REMOTE HOST IDENTIFICATION HAS CHANGED! @ @@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@ IT IS POSSIBLE THAT SOMEONE IS DOING SOMETHING NASTY! Someone could be eavesdropping on you right now (man-in-the-middle attack)! It is also possible that the RSA host key has just been changed. The fingerprint for the RSA key sent by the remote host is 5d:d2:f5:21:fa:07:64:0d:63:1b:3b:ee:a6:58:58:bb. Please contact your system administrator. Add correct host key in /root/.ssh/known_hosts to get rid of this message. Offending key in /root/.ssh/known_hosts:2 RSA host key for 192.168.1.102 has changed and you have requested strict checking. Host key verification failed. [root@bigboy tmp]#

If you are confident that the error is due to a reinstallation, then edit your ~/.ssh/known_hosts text file, removing the entry for the offending remote server. When you try connecting via SSH again, you’ll be prompted to add the new key to your ~/.ssh/known_hosts file and the login session should proceed as normal after that.

DEACTIVATING TELNET AFTER INSTALLING SSH
You should always consider SSH over TELNET, because of the inherent data encryption features of SSH and the current widespread availability of SSH clients for both Linux and Windows. By default, the TELNET server isn’t installed with Fedora Linux. If you do decide to deactivate an active TELNET server on Fedora, then use the chkconfig command as detailed in Chapter 16, “TELNET, TFTP, and xinetd.”
[root@bigboy tmp]# chkconfig telnet off

19Harrison_ch17.qxd

2/25/05

2:43 PM

Page 269

SCP: A Secure Alternative to FTP

269

EXECUTING REMOTE COMMANDS

ON

DEMAND WITH SSH

A nice feature of SSH is that it is capable of logging in and executing single commands on a remote system. You just have to place the remote command, enclosed in quotes, at the end of the ssh command of the local server. In the example below, a user on server Smallfry who needs to know the version of the kernel running on server Bigboy (192.168.1.100) remotely runs the uname -a command. The command returns the version of 2.6.8-1.521 and the server’s name, Bigboy:
[root@smallfry tmp]# ssh 192.168.1.100 “uname -a” root@192.168.1.100’s password: Linux bigboy 2.6.8-1.521 #1 Mon Aug 16 09:01:18 EDT 2004 i686 i686 i386 GNU/Linux [root@smallfry tmp]#

This feature can be very useful. You can combine it with password-free login, explained later in this chapter, to get the status of a remote server whenever you need it. More comprehensive monitoring may best be left to such purpose built programs as MRTG, which is covered in Chapter 22, “Monitoring Server Performance.”

SCP: A SECURE ALTERNATIVE TO FTP
From a networking perspective, FTP isn’t very secure, because usernames, passwords, and data are sent across the network unencrypted. More secure forms such as SFTP (Secure FTP) and SCP (Secure Copy) are available as a part of the OpenSSH package that is normally installed by default on Red Hat and Fedora Core. Remember, unlike FTP, SCP doesn’t support anonymous downloads like FTP. The Linux scp command for copying files has a format similar to that of the regular Linux cp command. The first argument is the source file and the second is the destination file. When copying to or from a remote server, SCP logs into the server to transfer the data and this therefore requires you to supply a remote server name, username, and password to successfully execute the command. The remote filename is therefore preceded by a prefix of the remote username and server name separated by an @ symbol. The remote filename or directory then follows separated by a colon. The format therefore looks like this:
username@servername:filename username@servername:directoryname

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 270

270

Secure Remote Logins and File Copying

Chapter 17

For example, file /etc/syslog.conf on a server with IP address 192.168.1.100 that needs to be retrieved as user peter would have the format peter@192.168.1.000:/etc/syslog.conf, the entire /etc directory would be
peter@192.168.1.000:/etc/.

Tip You can download an easy-to-use Windows SCP client called WinSCP from
http://winscp.vse.cz/eng/.

Copying Files to the Local Linux Box If you understand how scp represents remote filenames, you can start copying files fairly easily. For example, to copy file /tmp/software.rpm on the remote machine to the local directory /usr/rpm use the commands:
[root@bigboy tmp]# scp root@smallfry:/tmp/software.rpm /usr/rpm root@smallfry’s password: software.rpm 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

To copy the file /tmp/software.rpm on the remote machine to the local directory /usr/rpm using TCP port 435, use the commands:
[root@bigboy tmp]# scp -P 435 root@smallfry:/tmp/software.rpm /usr/rpm root@smallfry’s password: software.rpm 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

Copying Files to the Remote Linux Box Copying files to the local Linux server now becomes intuitive. For example, to copy file /etc/hosts on the local machine to directory /tmp on the remote server:
[root@bigboy tmp]# scp /etc/hosts root@192.168.1.103:/tmp root@192.168.1.103’s password: hosts 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

To copy file /etc/hosts on the local machine to directory remote server via TCP port 435, use the commands:

/tmp

on the

[root@bigboy tmp]# scp -P 435 /etc/hosts root@192.168.1.103:/tmp hosts 100% 1011 27.6KB/s 00:00 [root@bigboy tmp]#

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 271

Using SSH and SCP without a Password

271

SFTP: ANOTHER SECURE ALTERNATIVE TO FTP
OpenSSH also has the SFTP program, which runs over an encrypted SSH session but whose commands mimic those of FTP. SFTP can be more convenient to use than SCP when you are uncertain of the locations of the files you want to copy, because it has the directory browsing abilities of FTP. Unlike FTP, SFTP doesn’t support anonymous logins. Here is a sample login sequence that logs in, gets help on the available commands, and downloads a file to the local server:
[root@bigboy tmp]# sftp 192.168.1.200 Connecting to 192.168.1.200... root@192.168.1.200’s password: sftp> help Available commands: cd path Change remote directory to ‘path’ lcd path Change local directory to ‘path’ chgrp grp path Change group of file ‘path’ to ‘grp’ chmod mode path Change permissions of file ‘path’ to ‘mode’ ... ... sftp> ls .. .. anaconda-ks.cfg install.log install.log.syslog .. .. sftp> get install.log install.log 100% 17KB 39.4KB/s 00:00 sftp> exit [root@bigboy tmp]#

USING SSH

AND

SCP WITHOUT

A

PASSWORD

From time to time you may want to write scripts that will allow you to copy files to a server, or login, without being prompted for passwords. This can make them simpler to write and also prevents you from having to embed the password in your code. SCP has a feature that allows you to do this. You no longer have to worry about prying eyes seeing your passwords nor worry about your script breaking when someone changes the password. You can configure SSH to do this by generating and installing data transfer encryption keys that are tied to the IP addresses of the two servers. The servers then use these pre-installed keys to authenticate one another for each file transfer. As you may expect, this feature

19Harrison_ch17.qxd

2/25/05

2:43 PM

Page 272

272

Secure Remote Logins and File Copying

Chapter 17

doesn’t work well with computers with IP addresses that periodically change, such as those obtained via DHCP. There are some security risks though. The feature is automatically applied to SSH as well. Someone could use your account to log into the target server by entering the username alone. It is therefore best to implement this using unprivileged accounts on both the source and target servers. The example that follows enables this feature in one direction (from server Bigboy to server Smallfry) and only uses the unprivileged account called filecopy. Configuration: Client Side Here are the steps you need to do on the computer that acts as the SSH client: 1. Generate your SSH encryption key pair for the filecopy account. Press the Enter key each time you are prompted for a password to be associated with the keys. (Do not enter a password.)
[filecopy@bigboy filecopy]# ssh-keygen -t dsa Generating public/private dsa key pair. Enter file in which to save the key (/filecopy/.ssh/id_dsa): Enter passphrase (empty for no passphrase): Enter same passphrase again: Your identification has been saved in /filecopy/.ssh/id_dsa. Your public key has been saved in /filecopy/.ssh/id_dsa.pub. The key fingerprint is: 1e:73:59:96:25:93:3f:8b:50:39:81:9e:e3:4a:a8:aa filecopy@bigboy [filecopy@bigboy filecopy]#

2. These keyfiles are stored in the .ssh subdirectory of your home directory. View the contents of that directory. The file named id_dsa is your private key, and id_dsa.pub is the public key that you will be sharing with your target server. Versions other than Red Hat/Fedora may use different filenames, use the SSH man pages to verify this.
[filecopy@bigboy filecopy]# cd ~/.ssh [filecopy@bigboy filecopy]# ls id_dsa id_dsa.pub known_hosts [filecopy@bigboy .ssh]#

3. Copy only the public key to the home directory of the account to which you will be sending the file.
[filecopy@bigboy .ssh]# scp id_dsa.pub \ filecopy@smallfry:public-key.tmp

Now, on to the server side of the operation.

19Harrison_ch17.qxd

2/25/05

2:44 PM

Page 273

Conclusion

273

Configuration: Server Side Here are the steps you need to do on the computer that will act as the SSH server. 1. Log into Smallfry as user filecopy. Create an .ssh subdirectory in your home directory and then go to it with cd.
[filecopy@smallfry public-key.tmp [filecopy@smallfry [filecopy@smallfry [filecopy@smallfry filecopy]# ls filecopy]# mkdir .ssh filecopy]# chmod 700 .ssh filecopy]# cd .ssh

2. Append the public-key.tmp file to the end of the authorized_keys file using the >> append redirector with the cat command. The authorized_keys file contains a listing of all the public keys from machines that are allowed to connect to your Smallfry account without a password. Versions other than Red Hat/Fedora may use different filenames, use the SSH man pages to verify this.
[filecopy@smallfry .ssh]# cat ~/public-key.tmp >> authorized_keys [filecopy@smallfry .ssh]# rm ~/public-key.tmp

From now on you can use ssh and scp as user filecopy from server Bigboy to Smallfry without being prompted for a password.

CONCLUSION
Most Linux security books strongly recommend using SSH and SCP over TELNET and FTP because of their encryption capabilities. Despite this, there is still a place for FTP in the world thanks to its convenience in providing simple global access to files and TELNET, which is much easier to implement in price-sensitive network appliances than SSH. Consider all options when choosing your file transfer and remote login programs and select improved security whenever possible as the long-term benefits eventually outweigh the additional cost over time.

19Harrison_ch17.qxd

2/25/05

10:08 AM

Page 274

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 275

C H A P T E R

18

Configuring DNS

IN THIS CHAPTER
☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Introduction to DNS Downloading and Installing the BIND Packages Starting BIND The /etc/resolv.conf File Configuring a Caching Name Server Important File Locations Configuring a Regular Name Server Troubleshooting BIND Migrating Your Web Site In-House DHCP Considerations for DNS Simple DNS Security Conclusion

Domain Name System (DNS) converts the name of a Web site (www.linuxhometo an IP address (65.115.71.34). This step is important, because the IP address of a Web site’s server, not the Web site’s name, is used in routing traffic over the Internet. This chapter will explain how to configure your own DNS server to help guide Web surfers to your site.
networking.com)

INTRODUCTION TO DNS
Before you dig too deep in DNS, you need to understand a few foundation concepts on which the rest of the chapter will be built.

275

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 276

276

Configuring DNS

Chapter 18

DNS Domains Everyone in the world has a first name and a last, or family, name. The same thing is true in the DNS world: A family of Web sites can be loosely described as a domain. For example, the domain linuxhomenetworking.com has a number of children, such as www.linuxhomenetworking.com and mail.linuxhomenetworking.com for the Web and mail servers, respectively. BIND BIND is an acronym for the Berkeley Internet Name Domain Project, which is a group that maintains the DNS-related software suite that runs under Linux. The most well-known program in BIND is named, the daemon that responds to DNS queries from remote machines. DNS Clients A DNS client doesn’t store DNS information; it must always refer to a DNS server to get it. The only DNS configuration file for a DNS client is the /etc/resolv.conf file, which defines the IP address of the DNS server it should use. You shouldn’t need to configure any other files. You’ll become well acquainted with the /etc/resolv.conf file soon. Authoritative DNS Servers Authoritative servers provide the definitive information for your DNS domain, such as the names of servers and Web sites in it. They are the last word in information related to your domain. How DNS Servers Find Your Site Information There are 13 root authoritative DNS servers (super duper authorities) that all DNS servers query first. These root servers know all the authoritative DNS servers for all the main domains—.com, .net, and the rest. This layer of servers keep track of all the DNS servers that Web site systems administrators have assigned for their sub domains. For example, when you register your domain my-web-site.org, you are actually inserting a record on the .com DNS servers that point to the authoritative DNS servers you assigned for your domain. (More on how to register your site later.) When to Use a DNS Caching Name Server Most servers don’t ask authoritative servers for DNS directly, they usually ask a caching DNS server to do it on their behalf. The caching DNS servers then store (or cache) the most frequently requested information to reduce the lookup overhead of subsequent queries.

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 277

Introduction to DNS

277

If you want to advertise your Web site www.my-web-site.org to the rest of the world, then a regular DNS server is what you require. Setting up a caching DNS server is fairly straightforward and works whether or not your ISP provides you with a static or dynamic IP address. After you set up your caching DNS server, you must configure each of your home network PCs to use it as their DNS server. If your home PCs get their IP addresses using DHCP, then you have to configure your DHCP server to make it aware of the IP address of your new DNS server, so that the DHCP server can advertise the DNS server to its PC clients. Off-the-shelf router/firewall appliances used in most home networks usually can act as both the caching DNS and DHCP server, rendering a separate DNS server is unnecessary. You can find the configuration steps for a Linux DHCP server in Chapter 8, “Configuring the DHCP Server.” When to Use a Static DNS Server If your ISP provides you with a fixed or static IP address, and you want to host your own Web site, then a regular authoritative DNS server would be the way to go. A caching DNS name server is used as a reference only, regular name servers are used as the authoritative source of information for your Web site’s domain. Note Regular name servers are also caching name servers by default.

When To Use A Dynamic DNS Server If your ISP provides your router/firewall with its IP address using DHCP, then you must consider dynamic DNS covered in Chapter 19, “Dynamic DNS.” For now, I’m assuming that you are using static IP addresses. How to Get Your Own Domain Whether or not you use static or dynamic DNS, you need to register a domain. Dynamic DNS providers frequently offer you a subdomain of their own site, such as my-web-site.dnsprovider.com, in which you register your domain on their site. If you choose to create your very own domain, such as my-web-site.org, you have to register with a company specializing in static DNS registration and then point your registration record to the intended authoritative DNS for your domain. Popular domain registrars include VeriSign, Register Free, and Yahoo. If you want to use a dynamic DNS provider for your own domain, then you have to point your registration record to the DNS servers of your dynamic DNS provider. (More details on domain registration are coming later in the chapter.)

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 278

278

Configuring DNS

Chapter 18

Basic DNS Testing of DNS Resolution As you know, DNS resolution maps a fully qualified domain name (FQDN), such as www.linuxhomenetworking.com, to an IP address. This is also known as a forward lookup. The reverse is also true: By performing a reverse lookup, DNS can determining the fully qualified domain name associated with an IP address. Many different Web sites can map to a single IP address, but the reverse isn’t true; an IP address can map to only one FQDN. This means that forward and reverse entries frequently don’t match. The reverse DNS entries are usually the responsibility of the ISP hosting your site, so it is quite common for the reverse lookup to resolve to the ISP’s domain. This isn’t an important factor for most small sites, but some e-commerce applications require matching entries to operate correctly. You may have to ask your ISP to make a custom DNS change to correct this. There are a number of commands you can use to do these lookups. Linux uses the host command, for example, but Windows uses nslookup.
The host Command

The host command accepts arguments that are either the fully qualified domain name or the IP address of the server when providing results. To perform a forward lookup, use the syntax:
[root@bigboy tmp]# host www.linuxhomenetworking.com www.linuxhomenetworking.com has address 65.115.71.34 [root@bigboy tmp]#

To perform a reverse lookup:
[root@bigboy tmp]# host 65.115.71.34 34.71.115.65.in-addr.arpa domain name pointer 65-115-71-34.my-ispprovider.net. [root@bigboy tmp]#

As you can see, the forward and reverse entries don’t match. The reverse entry matches the entry of the ISP.
The nslookup Command

The nslookup command provides the same results on Windows PCs. To perform forward lookup, use:
C:\> nslookup www.linuxhomenetworking.com Server: 192-168-1-200.my-web-site.org Address: 192.168.1.200 Non-authoritative answer: Name: www.linuxhomenetworking.com Address: 65.115.71.34 C:\>

20Harrison_ch18.qxd

2/25/05

10:09 AM

Page 279

The /etc/resolv.conf File

279

To perform a reverse lookup:
C:\> nslookup 65.115.71.34 Server: 192-168-1-200.my-web-site.org Address: 192.168.1.200 Name: 65-115-71-34.my-isp-provider.net Address: 65.115.71.34 C:\>

DOWNLOADING

AND INSTALLING THE

BIND PACKAGES

Most Red Hat and Fedora Linux software products are available in the RPM format. When searching for the file, remember that the BIND RPM’s filename usually starts with the word “bind” followed by a version number, as in bind9.2.2.P3-9.i386.rpm. (For more details on downloading RPMs, see Chapter 6, “Installing RPM Software.”)

STARTING BIND
You can use the chkconfig command to get BIND configured to start at boot:
[root@bigboy tmp]# chkconfig named on

To start, stop, and restart BIND after booting, use:
[root@bigboy tmp]# service named start [root@bigboy tmp]# service named stop [root@bigboy tmp]# service named restart

Remember to restart the BIND process every time you make a change to the configuration file for the changes to take effect on the running process.

THE /ETC/RESOLV.CONF FILE
DNS clients (servers not running BIND) use the /etc/resolv.conf file to determine both the location of their DNS server and the domains to which they belong. The file generally has two columns; the first contains a keyword, and the second contains the desired values separated by commas. See Table 18.1 for a list of keywords.

org.100 and 192.168. therefore. DNS on your client appends the server name to each domain in this list and does a DNS lookup on each to get the remote servers’ IP addresses. which should be searched for shorthand references to other servers. There should be only one entry per nameserver keyword. you’ll have to comment out the reference to your previous DNS server (most likely your router) with a # or make it point to the server itself using the universal localhost IP address of 127.1.1. If you refer to another server just by its name without the domain added on.100 nameserver 192. Two name servers. Placing a domain and search entry in the /etc/resolv.168.org.1 .0. then the entry would just be my-web-site.1. So. The only file you have to edit is /etc/resolv.org nameserver 192. If there is more than one name server.168.org my-web-business.conf is redundant.1. use multiple nameserver lines.org and my-web-business. enabling you to refer to servers in the same domain by only their server name without having to specify the domain.org another-web-site.1. your old entry of: nameserver 192. 192. but it also is a member of domains another-website.1 Keyword Keywords in /etc/resolv.0. my-web-site. nameserver domain search Take a look at a sample configuration in which the client server’s main domain is my-web-site.1.org.org. The local domain name to be used by default. If the server is bigboy.org.168.conf Value IP address of your DNS name server. This is a handy time-saving feature. provide DNS name resolution: search my-web-site.qxd 2/25/05 10:09 AM Page 280 280 Configuring DNS Chapter 18 Table 18. CONFIGURING A CACHING NAME SERVER The Red Hat/Fedora default installation of BIND is configured to convert your Linux box into a caching name server. in this case my-web-site.1 would be replaced by a new entry of: # nameserver 192. The domains in this list must be separated by spaces.168.102 The first domain listed after the search directive must be the home domain of your network.168.1.conf.102.20Harrison_ch18.

2 provides a map. Sometimes BIND is also installed using Linux’s chroot feature to not only run named as user named. and the chroot BIND add-on RPM installs its own versions in their chroot locations.20Harrison_ch18. which returns the name of the RPM: [root@bigboy tmp]# rpm -q bind-chroot bind-chroot-9. the hacker’s access to the rest of your system is isolated to the files under the chroot directory and nothing else.conf /var/named/chroot/etc/ [root@bigboy tmp]# cp -f /etc/rndc. You can determine whether you have the chroot add-on RPM by using this command. the chroot versions of some of the files are empty. .0. IMPORTANT FILE LOCATIONS Red Hat/Fedora BIND normally runs as the named process owned by the unprivileged named user. named is fooled into thinking that the directory /var/named/chroot is actually the root or / directory. Therefore. it is important to understand exactly where the files are located. and those you’d expect to find in /var/named are actually located in /var/named/chroot/var/named. named files normally found in the /etc directory are found in /var/named/chroot/etc directory instead. The advantage of the chroot feature is that if a hacker enters your system via a BIND exploit.* /var/named/chroot/etc/ Before you go to the next step of configuring a regular name server. copy the configuration files to their chroot locations: [root@bigboy tmp]# cp -f /etc/named. but also to limit the files named can see.0.qxd 2/25/05 10:09 AM Page 281 Important File Locations 281 or nameserver 127.3-13 [root@bigboy tmp]# There can be confusion with the locations: Regular BIND installs its files in the normal locations. When installed.1 The next step is to make all the other machines on your network point to the caching DNS server as their primary DNS server. Unfortunately. Before starting Fedora BIND.2. This type of security is also known as a chroot jail. Table 18.

0. assume your ISP assigned you the subnet 97.my-web-site.zone.conf You’ll have to make your DNS server refer to itself for all DNS queries by configuring the /etc/resolv. Configuring resolv. The zone file is named my-site.255. nameserver 127.conf file contains the main DNS configuration and tells BIND where to find the configuration files for each domain you own. the file my-site.conf file to reference localhost only. CONFIGURING A REGULAR NAME SERVER For the purposes of this tutorial.key. Use the code: . This file usually has two zone areas: ☞ Forward zone file definitions list files to map domains to IP addresses ☞ Reverse zone file definitions list files to map IP addresses to domains In this example.0. rndc.158.2 File Differences in Fedora and Red Hat DNS File Locations Purpose Tells the names of the zone files to be used for each of your Web site domains Named authentication Link all the IP addresses in your domain to their corresponding servers BIND chroot Location /var/named/chroot/etc Regular BIND Location /etc named. you’ll set up the forward zone for www.24 with a subnet mask of 255. although not explicitly stated. Red Hat 9 and earlier versions don’t.1 Configuring named.20Harrison_ch18.253.org by placing entries at the bottom of the named.248 (/29).conf The named.conf rndc.qxd 2/25/05 2:45 PM Page 282 282 Configuring DNS Chapter 18 Table 18.zone should be located in the default directory of /var/named/chroot/var/named in a chroot or in /var/named in a regular one.255.conf /var/named/chroot/etc /var/named/chroot/var/named /etc /var/named zone files Note Fedora Core installs BIND chroot by default.conf file. and.

zone: zone “another-site. This isn’t important for the Windows clients on your network.zone”.1. you could modify the directive to: allow-query { 192.0/24. but some Linux applications require valid forward and reverse entries to operate correctly. file “my-site.zone for the 192. The forward domain lookup process for mysite.1. Reverse lookups operate similarly by scanning an IP address from left to right to get increasingly specific information about an address. Note The allow-query directive defines the networks that are allowed to query your DNS server for information on any zone.0/24 address space.1.168. your ISP handles the reverse zone entries for your public IP addresses.20Harrison_ch18. This reverse zone definition for named. you have to format entries to handle the reverse lookups for your IP addresses. to which all IP addresses belong.conf file where the main in-addr. notify no. Here is an example for another-site. and for reverse lookups the scan is from left to right.168. }.org” { type master. is followed by the first three octets of the IP address in reverse order.168.zone”.conf file to reference other Web domains you host. allow-query { any. This order is important to remember or else the configuration will fail. but the noticeable difference is that for forward lookups the scan is from right to left. file “another-site. Next.0/24 network: .qxd 2/25/05 10:09 AM Page 283 Configuring a Regular Name Server 283 zone “my-web-site. This difference can be seen in the formatting of the zone statement for a reverse zone in /etc/named. }. }.com” { type master. allow-query { any.168. notify no.com using a zone file named another-site. but you will have to create reverse zone entries for your SOHO/home environment using the 192.0 network. }. The similarity in both methods is that increasingly specific information is sought. For example.conf uses a reverse zone file named 192-168-1.1. to limit queries to only your 192. }. In addition. you can insert additional entries in the named.arpa domain. In most cases.com scans the FQDN from right to left to get increasingly more specific information about the authoritative servers to use.

In the absence of a suffix. They all have the general format: Name Class Type Data There are different types of records for mail (MX). The purpose of a TTL is to reduce the number of DNS queries the authoritative DNS server has to answer.196. notify no. ☞ By default. aliases (CNAME) and overall zone definitions. then caching servers use the original stored response for three days before making the query again: $TTL 3D BIND recognizes several suffixes for time-related values. BIND assumes the value is in seconds. file “192-168-1. They define the nature of the DNS information in your zone files that’s presented to querying DNS clients. forward lookups (A). DNS Resource Records The rest of the records in a zone file are usually BIND resource records. MX. If the TTL is set to three days. ☞ Each zone file contains a variety of records (SOA.zone”.qxd 2/25/05 10:09 AM Page 284 284 Configuring DNS Chapter 18 zone “1.in-addr. which is the period for which it’s valid.arpa” { type master. Take a closer look at these entries in the zone file. Configuring the Zone Files You need to keep a number of things in mind when configuring DNS zone files: ☞ In all zone files.20Harrison_ch18.168. Time to Live Value The very first entry in the zone file is usually the zone’s time to live (TTL) value. reverse lookups (PTR). and an H signifies hours. A D signifies days. A. and CNAME) that govern different areas of BIND. NS. a W signifies weeks. Caching DNS servers cache the responses to their queries from authoritative DNS servers. . }. you can place a comment at the end of any line by inserting a semi-colon character then typing in the text of your comment. your zone files are located in the directory /var/named or /var/named/chroot/var/named. The authoritative servers not only provide the DNS answer but also provide the information’s time to live.

and will sometimes wrap around on your screen. Home/SOHO will be limited to the IN or Internet class used when defining IP address mapping information for BIND. You can use the format YYYYMMDD with single digit incremented number tagged to the end to provide an incremental value that provides some editing information. A serial number for the current configuration. Table 18.3 explains what each field in the record means. 2004100801 . Here is an example: @ IN SOA ns1. Similarly. The SOA Record The first resource record is the SOA record. There are a number of different DNS classes. The data portion is formatted according to the record type and may consist of several values separated by spaces. The @ sign is a shorthand reference to the current origin (zone) in the /etc/named. Other classes exist for non-Internet protocols and functions but are very rarely used. The type of DNS resource record.org. hostmaster.3 Field Name Class The SOA Record Format Description The root name of the zone.my-web-site. Other types of records exist. you can insert new line characters between the fields as long as you insert parenthesis at the beginning and end of the insertion to alert BIND that part of the record will straddle multiple lines. Fully qualified name of your primary name server. The e-mail address must also be followed by a period. minimum Table 18. In the example. The regular @ in the e-mail address must be replaced with a period instead. The e-mail address of the name server administrator. Must be followed by a period. the name is also subject to interpretation based on this factor. You can also add comments to the end of each new line separated by a semicolon when you do this. this is an SOA resource record.20Harrison_ch18. It has the format: Name Class Type Name-Server Email-Address Serial-No Refresh Retry Expiry Minimum-TTL The record can be long. Type Name-server Email-address Serial-no continues . which I’ll cover later. 1H . expiry .org. ( serial # refresh retry 1W 1D ) . 4H .my-web-site. which contains general administrative and control information about the domain.qxd 2/25/05 10:09 AM Page 285 Configuring a Regular Name Server 285 Start of Authority (SOA).conf file for that particular database file. For the sake of formatting.

4 Record Type NS. 1 hour. and minimum values of 4 hours.my-web-site. Table 18. Table 18. Total amount of time a slave should retry to contact the master before expiring the data it contains. respectively. This value defines the caching duration your DNS includes in this response. Name of a server in the domain Server name alias Last octet of server’s IP address Class Field2 IN IN Type Field NS MX Data Field IP address or CNAME of the name server Mail server DNS name A CNAME PTR IN IN IN A CNAME PTR IP address of server “A” record name for the server Fully qualified server name 1 2 If the search key to a DNS resource record is blank. Slaves aren’t usually used in home/SOHO environments.qxd 2/25/05 10:09 AM Page 286 286 Configuring DNS Chapter 18 Table 18. The serial number is 2004100801 with refresh. Future references will be directed towards the root servers.20Harrison_ch18. Minimum-TTL So in the example. the NS. A. expiry. Slaves aren’t usually used in home/SOHO environments. retry. For most home/SOHO scenarios. You should also be aware that IN is the default Class. 1 week.org. it reuses the search key from the previous record. and CNAME records each occupy a single line with a very similar general format. and CNAME Record Formats Field Descriptions Name Field NS MX Usually blank1 Domain to be used for mail. PTR. A. The slave’s retry interval to connect the master in the event of a connection failure.3 Field Refresh Retry Expiry continued Description Tells the slave DNS server how often it should check the master DNS server. PTR. the primary name server is defined as ns1. NS. Your DNS server will respond with a no domain or NXDOMAIN response that the remote client caches.4 outlines the way they are laid out. the Class field will always be IN or Internet. .org with a contact e-mail address of hostmaster@my-web-site. usually the same as the domain of the zone file itself. MX. A. PTR. For the example. and BIND assumes a record is of this type unless otherwise stated. and 1 day. Slaves aren’t usually used in home/SOHO environments. There are times when remote clients will make queries for subdomains that don’t exist. MX. MX. and CNAME Records Like the SOA record. this is the SOA @ sign.

☞ The MX record for my-web-site.org .0. refresh.my-web-site. and you will find your mail server accepting mail only for the domain my-web-site. $TTL 3D @ IN SOA ns1. or CNAME record. Zone file for my-web-site. A. Inet Address of nameserver my-web-site.158.org points to the server named mail.com. The full zone file . Sample Forward Zone File Now that you know the key elements of a zone file. Primary name servers are more commonly called ns1 and secondary name servers ns2.20Harrison_ch18. retry. hostmaster. expire. seconds .qxd 2/25/05 10:09 AM Page 287 Configuring a Regular Name Server 287 If you don’t put a period at the end of a host name in a SOA. BIND assumes an A record with www refers to www.org.my-web-site.org is the name server for my-web-site. seconds 3600 . ( 200211152 . seconds 3600 ) .org. localhost bigboy mail ns1 www MX 10 mail .253. Primary Mail Exchanger A A CNAME CNAME CNAME 127.org.my-website. but if you forget to put the period after the domain in the MX record for my-web-site.0. So. In corporate environments there may be a separate name server for this purpose.my-web-site. .myweb-site.org at the end. it’s time to examine a working example for the domain my-web-site. therefore remote DNS caching servers will store learned DNS information from your zone for three days before flushing it out of their caches.org: .org .mysite. This may be acceptable in most cases. serial# 3600 . ☞ The minimum TTL value ($TTL) is three days. . minimum. BIND will automatically tack on the zone file’s domain name to the name of the host. NS www .org.26 bigboy bigboy bigboy Notice that in this example: ☞ Server ns1.org. .org.org. BIND attaches the my-web-site. NS. seconds 3600 .1 97.

1. 100 PTR bigboy.253.1. also requires valid reverse lookup capabilities. When DNS is set up in a redundant configuration. . seconds . Sample Reverse Zone File Now you need to make sure that you can do a host query on all your home network’s PCs and get their correct IP addresses. seconds minimum. All the entries in the first column refer to the last octet of the IP address for the network.158. 200303301 8H 2H 4W 1D ) . Also the PTR records cannot have CNAME aliases. could cause your slaves to have outdated information. So here you have an example of the name server.my-website. If they were all different machines. This is very important if you are running a mail server on your network.158u. and use the serial number to determine whether the data on the master has been updated.1. Filename: 192-168-1.x network.my-web-site.26 97. NS . www .my-web-site.org. This is an example of a zone file for the 192. NFS. seconds expire.20Harrison_ch18. which is used in network-based file access.zone . and Web server being the same machine.x . Notice how the main difference between forward and reverse zone files is that the reverse zone file only has PTR and NS records.org. seconds retry.org. . Zone file for 192. ( www. hostmaster. . Nameserver Address . even though the contents of the zone file have been modified. .168. so the IP address 192. .100 points to the name bigboy.134 97. because sendmail typically relays mail only from hosts whose IP addresses resolve correctly in DNS.org. . then you’d have an A record entry for each: www mail ns A A A 97. . Failing to increment the serial number. $TTL 3D @ IN SOA site.qxd 2/25/05 10:09 AM Page 288 288 Configuring DNS Chapter 18 ☞ ns1 and mail are actually CNAMEs or aliases for the Web server www. the slave DNS servers periodically poll the master server for updated zone file information.158.253.168.my-webserial number refresh.253.125 It is a required practice to increment your serial number whenever you edit your zone file.168. mail server.

you could have a reference to a zone file called my-site. Don’t worry.32 to 192. which are the addresses the DHCP server issues. SMTP mail relay wouldn’t work for PCs that get their IP addresses via DHCP if these lines weren’t included.my-web-site.158. ochorios.conf file in one of two views sections. dhcp-192-168-1-33. reggae.253.org.168.1.1.X with references to 192.253.org. Place your zone statements in the /etc/named. BIND has a way around this called views.zone file is fairly easy: Copy it from the my-site.1.253. The views feature allows you to force BIND to use predefined zone files for queries from certain subnets.org.my-web-site.158.158.36.168.my-web-site. Most home DSL sites wouldn’t qualify. .26 NAT IP address of your Web server.my-web-site. This entry would be inserted in the internal section. This means it’s possible to use one set of zone files for queries from the Internet and another set for queries from your home network.X network. What do the PCs on your home network need to see? They need to see DNS references to the real IP address of the Web server.my-web-site. which Internet users would see.zone for lookups by home users on the 192. I included entries for addresses 192.org.zone file and replace all references to 97.org.org. ISPs won’t usually delegate this ability for anyone with less than a Class C block of 256 IP addresses.my-web-site.253.qxd 2/25/05 10:09 AM Page 289 Configuring a Regular Name Server 289 103 102 105 32 33 34 35 36 PTR PTR PTR PTR PTR PTR PTR PTR smallfry. 192.158. dhcp-192-168-1-36. The second view called external lists the zone files to used for Internet users. dhcp-192-168-1-32. You could also have a file called my-site-home. Here’s a summary of how it’s done: 1.org.0 network.X. dhcp-192-168-1-35.168. Creating the my-site-home.conf entry would be inserted in the external section. You may also want to create a reverse zone file for the public NAT IP addresses for your home network.my-web-site.1. For example.1. Unfortunately. What You Need to Know About NAT and DNS The previous examples assume that the queries will be coming from the Internet with the zone files returning information related to the external 97. because NAT won’t work properly if a PC on your home network attempts to connect to the external 97. This /etc/named.100.my-web-site.168.org.zone for lookups related to the 97.26 address of the Web server.168. The first section is called internal and lists the zone files to be used by your internal network.20Harrison_ch18. dhcp-192-168-1-34.

0.” IN { type hint.ip6.in-addr. file “named.local”. Remember to increment your serial numbers! Here is a sample configuration snippet for the /etc/named.0. To do this. }. }.0.0. All the statements below were inserted after the options and controls sections in the file.0. // IPv6 Support zone “0. }. localhost. allow-update { none.127. . }.arpa” IN { type master. allow-update { none.0. zone “0. }.0.zone”. You must also tell the DNS server which addresses you feel are internal and external.0.127.0. file “localhost.20Harrison_ch18. view “internal” { // What the home network will see match-clients { localnets.0.0.ca”.arpa.0.0. }.0.conf file I use for my home network. }.0. zone “.0.0. Some built-in ACLs can save you time: ☞ ☞ ☞ Refers to the DNS server itself Refers to all the networks to which the DNS server is directly connected any: which is self explanatory localnets: localhost: Note You must place your localhost.0. file “named. file “named.0/24.0. // ACL statement acl “trusted-subnet” { 192.168.local”. you must first define the internal and external networks with access control lists (ACLs) and then refer to these lists within their respective view section with the match-clients statement.in-addr.0.0. “trusted-subnet”. 0. allow-update { none.0.0.qxd 2/25/05 10:09 AM Page 290 290 Configuring DNS Chapter 18 2.0.arp a” IN { type master.17.0.ip6.0.” zone statements in the internal views section.0.0.0. zone “localhost” IN { type master.0. }.0.0. and period “.

192. zone “my-web-site. allow-update { none. notify no. file “192-168-1. file “another-site.org” { type master. }. }.in-addr.com” { type master.arpa” IN { type master. zone “1. }. }. file “named. }. }.com” { type master. .in-addr.255” zone “255. }. file “another-site-home. }.20Harrison_ch18. zone “another-site. allow-query { any. // Prevents lookups for network addresses ending in “.168. zone “my-web-site. file “named.in-addr.zone”. allow-query { any. }. }. }.zone”. }. notify no. view “external” { // What the Internet will see match-clients { any. file “my-site-home. recursion no.org” { type master.broadcast”. }.zone”.zone”.arpa” IN { type master. allow-query { any.qxd 2/25/05 10:09 AM Page 291 Configuring a Regular Name Server 291 }. notify no. notify no. allow-update { none.0” zone “0. // Prevents lookups for broadcast addresses ending in “. allow-update { none. }. }. allow-query { any.arpa” IN { type master. file “my-site. }. }.zone”.zero”. zone “another-site.

the other trusted network (192.168.17. and localhost get DNS data from the zone files in the internal view. Some programs. Remember. this is purely an example. such as sendmail. but be sure to increment your configuration file serial number before doing this: [root@bigboy tmp]# service named restart Take a look at the end of your are no errors.” explains how to correctly configure your /etc/hosts file.local my-site. /var/log/messages file to make sure there Make Sure Your /etc/hosts File is Correctly Updated Chapter 3.1 -rw-r--r-. I included an ACL for network 192. The example home network doesn’t need to have the ACL statement at all as the built in ACLs localnets and localhost are sufficient.zone named.zone The configuration files above will not be loaded until you issue the proper command to restart the named process that controls DNS. require a correctly configured /etc/hosts file even though DNS is correctly configured.zone 2001 2001 2001 16:23 localhost.0 /24).ca named.1.local my-site.17.1 -rw-r--r-.0). trusted-subnet Loading Your New Configuration Files To load your new configuration files. first make sure your file permissions and ownership are okay in the /var/named directory: [root@bigboy [root@bigboy total 6 -rw-r--r-.zone named.1 -rw-r--r-.qxd 2/25/05 10:09 AM Page 292 292 Configuring DNS Chapter 18 In this example. So in this case the local network (192.1 [root@bigboy tmp]# cd /var/named named]# ll named named 195 Jul 3 named named 2769 Jul 3 named named 433 Jul 3 root root 763 Oct 2 named]# chown named * named]# chgrp named * named]# ll named named named named named named named named named]# 195 2769 433 763 Jul Jul Jul Oct 3 3 3 2 2001 2001 2001 16:23 localhost.20Harrison_ch18.0 /24 called to help clarify the use of ACLs in more complex environments.168.1 [root@bigboy [root@bigboy [root@bigboy total 6 -rw-r--r-.168. I then inserted a reference to the trusted-subnet in the match-clients statement in the internal view.1 -rw-r--r-. The network won’t need the trusted-subnet section in the match-clients line either. Once the ACL was defined. “Linux Networking.ca named. .1 -rw-r--r-.1 -rw-r--r-.

253.my-web-site.org to handle your domain. or you could give your Web server a second IP address using an IP alias. or whatever it is.26.my-web-site. If you only have one. which should be up to date. . you’ll have to take two steps: 1. “Linux Firewalls Using iptables. It normally takes about three to four days for your updated DNS information to be propagated to all 13 of the world’s root name servers.my-web-site. The actual IP address of the server is 192.158. You’ll therefore have to wait about this amount of time before starting to notice people hitting your new Web site.qxd 2/25/05 10:09 AM Page 293 Configuring a Regular Name Server 293 Configure Your Firewall The sample network assumes that the BIND name server and Apache Web server software run on the same machine protected by a router/firewall.168.26 in this case).20Harrison_ch18. This screen will prompt you for the server name only. but different name. so that at least one of the name servers is your new name server (97.158. you may want to consider taking a look at Chapter 14. (This screen prompts you for both the server’s IP address and name. and different name. You’ll have to use NAT for Internet users to be able to gain access to the server via the chosen public IP address. Fix Your Domain Registration Remember to edit your domain registration for my-web-site.org or www.1. which should have delayed values. plus a few well-known ones. Create a new name server record entry for the IP address 97. If your firewall is a Linux box. namely 97. which is a private IP address.) 2. Sometimes. the registrar requires at least two registered name servers per domain. create a second NAT entry on your firewall and then create the second name server record entry with the new IP address. Assign ns.253.100. then you could either create a second name server record entry with the same IP address. Once you’ve logged in with the registrar’s username and password. You’ll most likely want to test your new DNS server. usually provide a Web interface to help you manage your domain.org or whatever your name server is called. such as VeriSign and RegisterFree.253.org.” which describes how to do the network address translation and allow DNS traffic through to your name server. You can use the chapter’s troubleshooting section to test specific DNS servers for the information they have on your site.26 to map to ns.158. Domain registrars.

Copy the /etc/rndc.0. [root@bigboy tmp]# service named restart Stopping named: rndc: connect failed: connection refused .* /var/named/chroot/etc/ [root@bigboy tmp]# cp /etc/named.conf /var/named/chroot/etc/ 2. 127.32. Here’s a quick list of symptoms that indicate your files may not be located correctly. ☞ The named daemon starts without loading any zone files.* and /var/named/chroot/etc/ directory: /etc/named.conf [root@bigboy tmp]# cat /var/named/chroot/etc/rndc. 10.71#53 Nov 9 17:35:41 bigboy named[1157]: on 127.20Harrison_ch18.key [root@bigboy tmp]# To correctly relocate the files: 1.0.conf files to the [root@bigboy tmp]# cp -f /etc/rndc.1#53 Nov 9 17:35:41 bigboy named[1157]: face eth0.0. If you have the BIND chroot package installed. Here is a sample of the /var/log/messages file that shows that named doesn’t load them:. it should shutdown correctly without error the second time.3 -u using 1 CPU loading configuration listening on IPv4 interlistening on IPv4 intercommand channel listening command channel listening running ☞ Restarting named gives rndc “connect failed” messages: [root@bigboy tmp]# service named restart Stopping named: rndc: connect failed: connection refused [ OK ] Starting named: [ OK ] [root@bigboy tmp]# ☞ The chroot files are empty: [root@bigboy tmp]# cat /var/named/chroot/etc/named.qxd 2/25/05 10:09 AM Page 294 294 Configuring DNS Chapter 18 TROUBLESHOOTING BIND One of the most common culprits of BIND problems is incorrectly located chroot files.1#953 Nov 9 17:35:41 bigboy named[1157]: on ::1#953 Nov 9 17:35:41 bigboy named[1157]: starting BIND 9.41. make sure the configuration files are located in the chroot directory. Nov 9 17:35:41 bigboy named[1157]: named -t /var/named/chroot Nov 9 17:35:41 bigboy named[1157]: Nov 9 17:35:41 bigboy named[1157]: from ‘/etc/named.0.2. Restart named twice.conf’ Nov 9 17:35:41 bigboy named[1157]: face lo.

conf’ Feb 21 09:13:13 bigboy named[12026]: no IPv6 interfaces found Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface lo.0.0. (Linux logging is covered in Chapter 5.conf file if they fail to do so.inaddr.arpa/IN: loaded serial 51 . It is best to test this from both inside your network and from the Internet. Use it to make sure all your zone files are loaded when you start BIND/named.1#53 Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface wlan0.” Feb 21 09:13:13 bigboy named: named startup succeeded Feb 21 09:13:13 bigboy named[12026]: loading configuration from ‘/etc/named.arpa/IN: loaded serial 51 Feb 21 09:13:14 bigboy named[12026]: zone 1.inaddr.0.arpa/IN: loaded serial 1997022700 Feb 21 09:13:14 bigboy named[12026]: zone 1. NAT.100#53 Feb 21 09:13:13 bigboy named[12026]: listening on IPv4 interface eth0.arpa/IN: zone localhost/IN: loaded running startup succeeded General Troubleshooting Steps Once your files are in the right place.0.168. Check your /etc/named.1.in-addr.127. 192.qxd 2/25/05 10:09 AM Page 295 Troubleshooting BIND 295 [ OK ] Starting named: [ [root@bigboy tmp]# Stopping named: [ Starting named: [ [root@bigboy tmp]# OK ] service named restart OK ] OK ] 3.172. Check your /var/log/messages file for the loading of the zone files: Nov 9 loaded Nov 9 serial Nov 9 Nov 9 17:36:34 bigboy named[1180]: serial 1997022700 17:36:34 bigboy named[1180]: 42 17:36:34 bigboy named[1180]: 09:36:35 bigboy named: named zone 0.” 2. or port forwarding rules to your DNS server.192. “Troubleshooting with syslog.127.100#53 Feb 21 09:13:14 bigboy named[12026]: command channel listening on 127. Lack of connectivity could be caused by a firewall with incorrect permit.20Harrison_ch18. you can continue with the following troubleshooting steps: 1. Failure could also be caused by the named process being stopped.1#953 Feb 21 09:13:14 bigboy named[12026]: zone 0. Troubleshooting with TELNET is covered in Chapter 4.in-addr.0.0.16.16. “Simple Network Troubleshooting. Linux status messages are logged to the file /var/log/messages.1. 127. 172. Determine whether your DNS server is accessible on DNS UDP/TCP port 53.168.

[root@bigboy tmp]# host www.linuxhomenetworking.71.168. ☞ Use the dig command to determine whether the name server for your domain is configured correctly.my-web-site.my-web-site.com www.linuxhomenetworking.org for the IP address of www.org Address: 192.) [root@bigboy tmp]# host www.com. try: ☞ Double check for your updated serial numbers in the modified files and also inspect the individual records within the files for mistakes.org/IN: running 3.1. Failure in this case could be due not only to an error on your BIND configuration or domain registration but also to an error in your DNS client’s DNS server entry in your Linux /etc/resolv.34 [root@bigboy tmp]# Here is an example of querying your default DNS server for the IP address of www.my-site.20Harrison_ch18.115.linuxhomenetworking.com Using domain server: Name: ns1.com.) . (You can also replace the name server’s name with its IP address.com has address 65.34 [root@bigboy tmp]# ☞ You can also use the dig command to determine whether known DNS servers on the Internet have received a valid update for your zone. Use the host (nslookup in Windows) command for both forward and reverse lookups to make sure the zone files were configured correctly.115.100#53 Aliases: www. Here is an example of querying DNS server ns1. the name of the specific DNS server to query has been left off the end.conf file or the Windows TCP/IP properties for your NIC.qxd 2/25/05 10:09 AM Page 296 296 Configuring DNS Chapter 18 Feb 21 loaded Feb 21 serial Feb 21 loaded Feb 21 09:13:14 bigboy named[12026]: serial 2004021401 09:13:14 bigboy named[12026]: 42 09:13:14 bigboy named[12026]: serial 200301114 09:13:14 bigboy named[12026]: zone my-web-site.com has address 65.com ns1. If this fails. ☞ Ensure there isn’t a firewall that could be blocking DNS traffic on TCP and/or UDP port 53 between your server and the DNS server.linuxhomenetworking.org/IN: zone localhost/IN: loaded zone my-web-site.linuxhomenetworking.71. (Remember if you decide to change the DNS servers for your domain that it could take up to four days for it to propagate across the Internet.linuxhomenetworking. As you can see.

115.my-isp-provider... [root@bigboy tmp]# dig linuxhomenetworking.. .com . 3600 IN NS ns2.69 Here is a successful dig using DNS server ns1.my-ispprovider.. 3600 IN NS ns2.conf file. .net.com linuxhomenetworking.net.com.70. [root@bigboy tmp]# 3600 3600 IN IN A A 65.68 65.. It is sometimes good to query both your name server.70.com SOA ..net. ns2.70.yahoo.. AUTHORITY SECTION: linuxhomenetworking.yahoo. . ADDITIONAL SECTION: ns1. it returns the SOA record for the zone. .com SOA .com... as well as a well-known name server such as ns1.com domain at VeriSign.115.115.115. ns1.69 Sometimes your SOA dig will fail.my-ispprovider. .. because it doesn’t refer to the /etc/resolv.20Harrison_ch18..yahoo.. In this case the authority section doesn’t know of the domain and points to the name server for the entire . This command uses the DNS server for the query. linuxhomenetworking.net..my-isp-provider. [root@bigboy tmp]# dig ns1. It returns the SOA record information and the addresses of the domain’s DNS servers in the authority section.. If you specify a name server.70.my-isp-provider. 3600 IN NS ns1. .net.com.net.my-isp-provider.my-isp-provider.68 65.yahoo..com to make sure your DNS records have propagated properly. .net.com for the query. 3600 IN NS ns1. . ADDITIONAL SECTION: ns1.[root@bigboy tmp]# 3600 3600 IN IN A A 65.com. This command uses the local DNS server for the query.qxd 2/25/05 10:09 AM Page 297 Troubleshooting BIND 297 The format for the command is: dig <domain-name> <name-server> soa The name server is optional. ns2. linuxhomenetworking. As before. ... AUTHORITY SECTION: linuxhomenetworking.. The dig command only works with fully qualified domain names. .my-isp-provider.. then dig queries that name server instead of the Linux server’s default name server.net..

but there is a lag in the propagation of the domain information across the Internet. ns1... QUESTION SECTION: . IN SOA . you can then revert to the old configuration. . As the TTL is usually set to a number of days.linuxhomeqnetworking. therefore the test server will begin to think that www..com.net. This file is usually given a higher priority than DNS.org refer to its own IP address. nstld. it will take at least three to five days for all remote DNS servers to recognize the change. AUTHORITY SECTION: com. 1077341254 1800 900 604800 900 . ☞ A firewall could be blocking DNS traffic on TCP and/or UDP port 53 between your server and the DNS server.org site that is currently in production.. Delays of up to four days are not uncommon. 2. not that of the www. There is no magic bullet that will allow you to tell all the caching DNS servers in the world to flush their caches of your zone file entries.. say one minute. . [root@bigboy tmp]# Possible causes of failure include: ☞ Typographical errors. .org if the new Web server is going to also be your new mail server.. If anything goes wrong.my-web-site.myweb-site.my-website.org is really hosted on itself. Edit the /etc/hosts file to make www. MIGRATING YOUR WEB SITE IN-HOUSE It is important to have a detailed migration plan if you currently use an external company to host your Web site and wish to move the site to a server at home or in your office.org in the DNS zone file to a very low value. 0 IN SOA a. You may also want to add an entry for mail.my-web-site.qxd 2/25/05 10:09 AM Page 298 298 Configuring DNS Chapter 18 [root@bigboy tmp]# dig .20Harrison_ch18.gtld-servers. Set up your test server in-house. At the very least your plan should include these steps: 1.. Once the propagation is complete.. In this case the misspelling “linuxhomeqnetworking. knowing it will rapidly recover within minutes rather than days. it will take only one minute to see the results of the final DNS configuration switch to your new server.com SOA .verisign-grs.yahoo.com.com” was entered on the command line..com linuxhomeqnetworking.. Your best alternative is to request your existing service provider to set the TTL on my-web-site. ☞ Incorrect domain registration. ☞ Correct domain registration.

org to point to your new Web server. Your client will usually refer to these files first before checking DNS. Fix your /etc/hosts files by deleting the test entries you had before. . this could be left in the hands of your service provider. If your Linux box is the DHCP server. Once testing is completed. Once complete. As the TTLs were set to one minute previously. you’ll be able to see results of the migration within minutes. and so on. Edit your my-web-site.my-web-site. Remember. You may also want to add an entry for mail. DHCP CONSIDERATIONS FOR DNS If you have a DHCP server on your network. 8. you’ll need to make it assign the IP address of the Linux box as the DNS server it tells the DHCP clients to use. 7.qxd 2/25/05 10:09 AM Page 299 Simple DNS Security 299 3. then you may need to refer to Chapter 8. You can then migrate these services inhouse as your confidence in hosting becomes greater. “Configuring the DHCP Server.org map to the IP address of the new server. RegisterFree or whoever you bought your domain from to point to your new DNS servers. You may also want to take over your own DNS.org if the new Web server is going to also be your new mail server. if you have concerns that your service provider won’t cooperate. You should take some precautions to conceal some of the information for the sake of security.org DNS entries with VeriSign. then you could explain to the provider that you want to test its failover capabilities to a duplicate server that you host in-house. the file would be C:\WINDOWS\system32\drivers\etc.” SIMPLE DNS SECURITY DNS can reveal a lot about the nature of your domain.my-website. Test the server from a remote client. Finally. coordinate with your Web hosting provider to update your domain registration’s DNS records for www. you can set the TTL back to the original value to help reduce the volume of DNS query traffic hitting your DNS server.my-web-site. You can test the server running as www. 6. 5. you don’t have to host DNS or mail in-house.20Harrison_ch18. 4. Test your server-based applications from the server itself. Just edit your /etc/hosts file on your Web browsing Linux PC to make www. hence you can use them to predefine some DNS lookups at the local client level only. You can then decide whether the change will be permanent once you have failed over back and forth a few times. Web. This should include mail. In the case of Windows.org even though DNS hasn’t been updated.my-web-site.

XFR size: 16 records [root@smallfry tmp]# This may not seem like an important security threat at first glance.1 mail.100) .org. 2004110701 3600 3600 3600 3600 .168. 3600 IN CNAME www.org.1.my-website.my-web-site. but the dig command is much more powerful.my-web-site. zone transfers should be disabled. 3600 IN SOA www.my-web-site.96 192-168-1-97.20Harrison_ch18. 3600 IN MX 10 mail.my-web-site.my-website. 3600 IN A 192.org.org. 3600 IN CNAME www.168.168. 3600 IN CNAME www. Anyone can use this command to determine all your server’s IP addresses and from the names determine what type of server it is and then launch an appropriate cyber attack.1.org.100 gateway.1. ntp.97 192-168-1-98. hostmaster.my-web-site.1.my-web-site. .org.org AXFR .1. 3600 IN A 192.org zone file: [root@smallfry tmp]# dig my-web-site.org.100#53(192.168.my-website. my-web-site.my-web-site. 3600 IN SOA www. but it is.org.org.168. 3600 IN A 192. ns1.}.org.my-web-site.my-website. smallfry.. You can do this by applying the allow-transfer directive to the global options section of your named. <<>> DiG 9. 3600 IN A 192.2. Query time: 16 msec .my-web-site. 3600 IN A 127. 3600 IN NS ns1.org.168.1.my-website.conf file.org. 3600 IN A 192.org.168.org..168.0.3 <<>> my-web-site.org.org.my-website. In this example.my-web-site. options { allow-transfer {none.org.org. SERVER: 192. WHEN: Sun Nov 14 20:21:07 2004 .my-website.qxd 2/25/05 10:09 AM Page 300 300 Configuring DNS Chapter 18 Zone Transfer Protection The host command does one DNS query at a time. 192-168-1-96.1..168.my-web-site.org.98 bigboy. }.1.. hostmaster.0.1.org AXFR . global options: printcmd my-web-site..1 localhost.my-web-site.org. 3600 IN A 192.org. In a simple home network. 3600 IN A 192.100 my-web-site.102 www. When given the right parameters it can download the entire contents of your domain’s zone file.org. the AFXR zone transfer parameter is used to get the contents of the my-web-site. 2004110701 3600 3600 3600 3600 my-web-site. without master and slave servers.org.my-web-site.org.

and they should remain obvious to all. [root@smallfry tmp]# Naming Convention Security Your my-web-site. [root@smallfry tmp]# dig my-web-site. <<>> DiG 9. . your MySQL database server doesn’t have the letters “DB” or “SQL” in the name. Dynamic DNS was created as a solution to this and is explained in Chapter 19. this is just one of many methods you can use. but to the Internet these names provide rapid identification of the types of malicious exploits a hacker could use to break in. DNS modifications are usually infrequent. although they can be a little complicated. CONCLUSION DNS management is a critical part of the maintenance of any Web site. your zone transfer test should fail.” ..20Harrison_ch18. or that your firewall doesn’t have the letters “FW” in its name either. You may want to adjust your DNS views so that to external users.3 <<>> my-web-site. global options: printcmd .. making DNS management extremely difficult.qxd 2/25/05 10:09 AM Page 301 Conclusion 301 Once applied.org AXFR . This may be good for ease of reference within the company. There are situations in which a server’s IP address will change unpredictably and frequently.org domain will probably have a www and a mail subdomain.2. because the IP address of a server is normally fixed or static... “Dynamic DNS..org AXFR . This is not always the case. Transfer failed. Fortunately. . Web site security refers to anything that helps to guarantee the availability of the site.

qxd 2/25/05 10:09 AM Page 302 .20Harrison_ch18.

This chapter will explain the details of dynamic DNS configuration. ☞ Dynamic DNS (DDNS): Used when you get a changing dynamic IP addresses via DHCP from your ISP.org domain. For this reason there are two broad types of DNS: ☞ Static DNS: This is used when your ISP provides you with unchanging fixed or static IP addresses.” which assume the IP address of a server won’t change continuously. Your DNS server acts as the authoritative source of information for your my-web-site. This can cause problems with the DNS zone files explained in Chapter 18.21Harrison_ch19. the DSL or cable modem IP address is provided by DHCP and therefore changes from time to time.org domain. You can consider static DNS as the “traditional” or “regular” form of DNS.qxd 2/25/05 10:09 AM Page 303 C H A P T E R 19 Dynamic DNS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ Dynamic DNS Preparation Dynamic DNS and NAT Router/Firewalls DDNS Client Software: SOHO Router/Firewalls DDNS Client Software: Linux DDclient Testing Your Dynamic DNS Conclusion In many home networking environments. 303 . You will have to use the services of a third-party DNS provider to provide DNS information for your my-web-site. “Configuring DNS.

Setup You need to do some basic setup before testing can begin. Ask a friend somewhere else on the Internet to try to TELNET to this IP address on port 80. Use the ipconfig command to determine the new IP address of your Web server. Make sure you can ping your default gateway. Although this works well. it has disadvantages.qxd 2/25/05 10:09 AM Page 304 304 Dynamic DNS Chapter 19 DYNAMIC DNS PREPARATION Unlike DSL. The cost of the service can make hosting with a $10 /month virtual hosting service look very attractive. 4. If port 80 works. 1. and configure the Ethernet NIC for DHCP as described in Chapter 3.” If you can get through to the IP address on port 80. “The Apache Web Server.) 5. then you most likely have Apache configured correctly Testing from the Internet You may be able to see Web pages from the Web server itself.” 3. Connect your Linux server directly to your cable or DSL modem. Under the system. (This command is explained in Chapter 3 also. Testing ISP Connectivity for Your Web Site The very first thing you need to do is to determine whether your ISP allows inbound connections on your DSL or cable modem line. Web queries first hit their servers on the regular TCP ports and then these servers automatically redirect the Web clients to use the IP address of your server on a different TCP port. then ask the friend to point a Web browser to the IP address and see whether a valid Web page appears. Here is how you can test it out for yourself. and many search engines do not index redirected pages. but in some cases they’ll say no when in fact the answer is yes.” 2. If your Linux server will eventually be placed behind a firewall. Configure and start Apache on your Linux Web server as described in Chapter 20. . try to TELNET to this IP address on port 80 as explained in Chapter 4. “Linux Networking. Many DDNS providers are aware of this and provide a redirect service to bypass the problem.21Harrison_ch19. “Simple Network Troubleshooting. From the Linux Web server itself. most cable modem providers may not allow you to host sites at home by blocking inbound HTTP (TCP port 80) and SMTP mail (TCP port 25) while allowing most other TCP traffic through. The easiest way to do this is to phone them and ask. then adjust your network topology accordingly and test port 80 port forwarding to your Web server.

You can determine the external IP address of your router or firewall by logging onto your Linux Web server and issuing the curl command to query the DynDNS. your ISP probably doesn’t allow HTTP access to its networks. Configure your Web server to run on a different TCP port. Listen 80 Test again with TELNET on this new port. If it works. change “80” to some other value. The Apache httpd. If the test port is 1234.conf file uses the listen directive to do this. you need to choose between the broad categories of Dynamic DNS service: . you are interested in only the line that tells you what the server thinks your IP address is.qxd 2/25/05 10:09 AM Page 305 Dynamic DNS Preparation 305 If the TELNET to port 80 fails. then you can additionally sign up for a redirect service with your DDNS provider as explained earlier. use the grep command to extract just that part.org IP information server. In this case.97. the IP address is 24.org/ | grep -i address <html><head><title>Current IP Check</title></head><body>Current IP Address: 24. In this example.4. If you can get a connection with correctly displayed pages on a nonstandard port.4.conf listen directive. Let Apache run on port 80.dyndns. One of the features of the curl troubleshooting utility is that it allows you to download a Web page’s HTML code to your screen without interpreting it.110: [root@bigboy tmp]# curl -s http://checkip. then adjust your network topology accordingly.97. preferably above 1024.110</body></html> [root@bigboy tmp]# Registering DDNS Once you decide to go ahead with DDNS. try the Web browser test too. and test port forwarding from the nonstandard port to port 80 on your Web server from the Internet. then use a URL of http://server-ipaddress:1234. Note If you are running iptables. Test Port Forwarding If your Linux server will eventually be placed behind a firewall. remember to adjust the rules to match this new port or stop iptables temporarily while doing this testing.21Harrison_ch19. Change it to your new value and restart httpd: # httpd.

my-web-site. follow these steps: 1. 2.my-website. Another disadvantage is that you must rely on your DDNS provider staying in business or else you may lose your domain. ☞ Paid Customized DNS: You can register the domain name of your choice and still host your Web site on a DHCP line. If you choose to create your own domain and use a paid DDNS service. and register your Web sites (sometimes called hosts) as part of your domain (as in www. Create an account with the DDNS provider. which you’ll need to use when configuring your DDNS client.my-isp.qxd 2/25/05 10:09 AM Page 306 306 Dynamic DNS Chapter 19 ☞ Free Dynamic DNS: Your Web site name will be a subdomain of the DDNS provider’s domain. now comes in a RPM format.21Harrison_ch19.net. for example.my-website. Update your domain information with your main DNS registrar (Verisign and RegisterFree) to tell them to direct queries to *. 3.net. Install a DDNS client on your Web servers that continuously runs. Register your domains (such as my-web-site. The very popular one DDclient.org to the DNS name servers of the DDNS provider. then your site will become my-site. You should also be prepared for slower response times for your homebased site than if you were using a static IP and a regular DNS service. Install a DDNS Client on Your Server All DDNS service providers require that you use a DDNS client on your Web server that will periodically update the IP address information in your provider’s DDNS record. only updating the DDNS provider’s DNS servers with the most current DHCP IP address of the site whenever it detects a change.org) with them. Your DDNS registration process will provide you with a username and password.org) with such companies as Verisign and RegisterFree. For example. When you own your own domain this worry largely goes away as you are fully in control of the DNS registration and renewal process.org and mail. Remember that this type of service may be undesirable for a company that wants to establish its own corporate identity. . You can perform all the necessary steps on your DDNS service provider’s Web site. 4. if the DDNS provider’s domain is isp.

If the Linux box is being protected behind a NAT router/firewall. The RPM name usually starts with ddclient followed by a version number.rpm. there is no need to install software on your Web server. most home router/firewalls use network address translation (NAT) to map a single.qxd 2/25/05 10:09 AM Page 307 DDNS Client Software: Linux DDclient 307 DYNAMIC DNS AND NAT ROUTER/FIREWALLS As discussed in Chapter 2. or you can find it at rpmfind. “Linux Firewalls Using iptables. If your firewall is Linux based. DDNS CLIENT SOFTWARE: LINUX DDCLIENT One of the most commonly used clients is DDclient.” will be helpful. Many Web-based small office/home office (SOHO) firewalls have easy interfaces to configure port forwarding. as in ddclient-3. (If you need a refresher on RPMs. public. With this support.” to conserve the limited number of IP addresses available for Internet purposes. Like most Red Hat and Fedora Linux software products. then the examples in Chapter 14. DDNS CLIENT SOFTWARE: SOHO ROUTER/FIREWALLS Most new SOHO router/firewalls have built in dynamic DNS clients for one or more of the major DDNS service providers.noarch. Some DDNS providers use more intelligent clients. The reported value is therefore invalid. which can overcome the NAT limitations of DDNS by actually logging into your SOHO firewall to determine the latest IP address information.6. “Introduction to Networking. that can be configured to let the DDNS provider record the public IP address from which the data stream is originating. Once this is done.”) You can usually download the software from your DDNS provider. In these cases. please refer to your manufacturer’s manual on how to do this. DDclient is available in the RPM format. NAT can fool the operation of some DDNS client software.3-1. DHCP-obtained IP address to the many private IP addresses within your network. the software can report only the true IP address of the Linux box’s NIC interface. There is usually a Dynamic DNS Web menu that prompts for the name of the service provider and your DDNS username and password. “Installing RPM Software. consult Chapter 6. you’ll have to also configure your router/firewall to do port forwarding to make all HTTP traffic destined for the IP address of the router/firewall to be exclusively address translated using NAT and forwarded to a single server on your home network.21Harrison_ch19. then the NIC reports in its data stream to the DDNS provider a private IP address that no one can reach directly via the Internet.net. such as DDclient. .

conf file is usually installed completely commented out. only the use=web option gives a valid Internet IP address and should be considered as a first option: [root@bigboy tmp]# ddclient -daemon=0 –query use=if. if=lo address is 127. The most important general parameters to configure are: ☞ Username: Your DDNS account’s login name ☞ Password: Your DDNS account’s password ☞ Use: The method used to determine the IP address to advertise to the DDNS server Some important DDNS provider parameters are: ☞ Server: The name of the DDNS provider’s main DNS server ☞ Protocol: The methodology the DDNS client should use to communicate with the DDNS server ☞ Your domain: The domain to which your Web server will belong You can use the ddclient command to determine the best use parameter to use in the ddclient.21Harrison_ch19.1 use=if.dnsalias. protocol=dyndns2 mysite-example.dnsalias.26 [root@bigboy tmp]# Take a look at a sample configuration.168.com \ \ You can add one of the following use lines to the General Parameter Section near the top of the file to define the method that will be used to determine the correct IP address: .conf File The ddclient.0. In this example.com. This example specifies a username of my-account-login-name and a password of my-account-password using the dyndns DDNS service provider’s settings to track the Web site named mysite-example.253.100 use=web. if=wlan0 address is 192.conf file. web=dyndns address is 97. but provides many configuration examples for the most popular DDNS providers.qxd 2/25/05 10:09 AM Page 308 308 Dynamic DNS Chapter 19 The /etc/ddclient. # General Parameter Section login=my-account-login-name password=my-account-password # DDNS Provider Parameters Section server=members.1.0.dyndns.158.org.

This method is the simplest as it requires no further information and handles NAT correctly. stop. This option is good for NAT environments where the use=web option isn’t considered a good alternative.qxd 2/25/05 10:09 AM Page 309 DDNS Client Software: Linux DDclient 309 ☞ Query a well-known Internet server: The Web method queries two well known servers run by DynDNS. you can create your own parameters as outlined in the DDclient read-me file. use=web ☞ Use the IP address of a specific server NIC: You can also query the IP address of the DDclient Web server’s NIC interface of your choice. if=eth0 ☞ Login to your SOHO firewall for information: The ddclient. you should get a response of plain old process ID numbers: [root@bigboy tmp]# pgrep ddclient . This is probably most valuable for servers connected directly to the Internet. as in: [root@bigboy tmp]# service ddclient start [root@bigboy tmp]# service ddclient stop [root@bigboy tmp]# service ddclient restart Remember to restart the DDclient process every time you make a change to the ddclient. How to Get DDclient Started You can configure DDclient to start at boot time using the chkconfig command: [root@bigboy tmp]# chkconfig ddclient on You can start. If your model isn’t listed.org and DNSpark to determine the public Internet IP address of the Web server running the DDclient software.conf file for the changes to take effect on the running process. and not via NAT. you’ll have to start ddclient as shown in the next section. and restart DDclient after boot time using the DDclient initialization script.21Harrison_ch19. You can test whether the DDclient process is running with the pgrep command. use=if.conf file has a list of use statements for various vendor’s firewalls. After editing your configuration file.

6. If you failed to add your host record. you will get an error message like: [root@bigboy tmp]# host www.cisco [root@bigboy tmp]# locate -u If the command doesn’t work..3/README. ☞ Using the host www.qxd 2/25/05 10:09 AM Page 310 310 Dynamic DNS Chapter 19 Finding DDclient Help The DDclient read-me and ddclient.0.1#53 ** server can’t find www.21Harrison_ch19.6.. [root@bigboy tmp]# Alternately.my-web-site.org: NXDOMAIN .org command from your Linux command prompt to determine whether you are getting a valid response. /usr/doc/ddclient-3. you can use the files: locate command to find all the DDclient [root@bigboy tmp]# locate ddclient | grep READ /usr/doc/ddclient-3. try updating your locate database with the command followed by the locate ddclient command once more: [root@bigboy tmp]# locate -u TESTING YOUR DYNAMIC DNS You can test your dynamic DNS by: ☞ Looking at the status page of your DNS provider and making sure the IP address that matches your www site is the same as your router/firewall’s public IP address.1 Address: 127. Use the rpm command to get a list of installed DDclient files.0.3/README /usr/doc/ddclient-3.0.conf files are good sources of information for doing custom configurations and troubleshooting.3/README .my-web-site..my-web-site. one of which is the read-me file: [root@bigboy tmp]# rpm -ql ddclient . You can find the read-me file by using DDclient RPMs or the locate command.org Server: 127.0..6.

It is a good starting place to help you become familiar with Web hosting.my-web-site.net www.my-web-site.org has address 97. or managed as closely as business class lines. You can test to make sure everything is okay by forcing NS lookup to query the name servers directly.org ns1.21Harrison_ch19.net: [root@bigboy tmp]# host www.26 [root@bigboy tmp]# CONCLUSION Always remember that dynamic DNS works. but it is frequently unreliable as residential class broadband data circuits are not monitored.158. The example below queries the miniDNS name server ns1.minidns. maintained.253. . you may need to consider a regular data center far away from spilt coffee and the washing machine that always trips the circuit breakers.qxd 2/25/05 10:09 AM Page 311 Conclusion 311 This error could be because your domain hasn’t propagated fully throughout the Internet. but as your Web site becomes busier and more financially important to you.minidns.

21Harrison_ch19.qxd 2/25/05 10:09 AM Page 312 .

This chapter explains how to configure Apache in a number of commonly encountered scenarios for small Web sites.22Harrison_ch20. remember that the Apache RPM’s filename usually starts with the word httpd followed by a version number. as in httpd-2. When searching for the file.qxd 2/25/05 10:11 AM Page 313 C H A P T E R 20 The Apache Web Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Downloading and Installing the Apache Package Starting Apache Configuring DNS for Apache DHCP and Apache General Configuration Steps Configuration: Multiple Sites and IP Addresses Using Data Compression on Web Pages Apache Running on a Server Behind a NAT Firewall Protecting Web Page Directories with Passwords The /etc/httpd/conf.” 313 .d Directory Troubleshooting Apache Conclusion Apache is probably the most popular Linux-based Web server application in use. “Installing RPM Software. DOWNLOADING AND INSTALLING THE APACHE PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format. you’ll need to configure Apache to accept surfers wanting to access your Web site.2.0. Once you have DNS correctly set up and your server has access to the Internet. see Chapter 6. It is best to use the latest version of Apache.48-1.rpm. (For more on RPMs.

then you need to use dynamic DNS to get the correct Internet DNS entry for your Web server. “Configuring DNS. To do this.168. not dynamic DNS. CONFIGURING DNS FOR APACHE Remember that you will never receive the correct traffic unless you configure DNS for your domain to make your new Linux box Web server the target of the DNS domain’s www entry.0 network uses a fixed.22Harrison_ch20. If your Web server and firewall are different machines. This could cause your firewall port forwarding. as well.” DHCP AND APACHE As you remember. or static IP address that is outside of the range of the DHCP server to prevent you from having this problem. if your Internet connection uses DHCP to get its IP address. Remember that the IP address of your Web server can change if it gets its IP address using DHCP.d directory to start.) DHCP on your protected home network is different. The firewall uses NAT and port forwarding to pass Internet traffic on to the Web server. refer to Chapter 18. the Web server lives on the 192. stop.168.qxd 2/25/05 10:11 AM Page 314 314 The Apache Web Server Chapter 20 STARTING APACHE Use the chkconfig command to configure Apache to start at boot: [root@bigboy tmp]# chkconfig httpd on Use the httpd init script in the restart Apache after booting: /etc/init.1.1. “Dynamic DNS. In this case. and [root@bigboy tmp]# service httpd start [root@bigboy tmp]# service httpd stop [root@bigboy tmp]# service httpd restart You can test whether the Apache process is running with: [root@bigboy tmp]# pgrep httpd You should get a response of plain old process ID numbers. then you probably also need to set up port forwarding for your Web traffic to reach the Web server correctly. to break. (Chapter 19 explains port forwarding.0 home network protected by a firewall. I recommend that your Web server on the 192. In the book’s sample topology. .” or Chapter 19.

html. or directives. Where to Put Your Web Pages All the statements that define the features of each Web site are grouped together inside their own <VirtualHost> section.html.org with a DocumentRoot directory of /home/www/site1/.* 1 root root 48590 Jun 18 23:43 index. inside a <VirtualHost> container are: ☞ servername: Defines the name of the <VirtualHost> container. The most commonly used statements. as I’ll explain soon.html pointing to the file index. page named index.my-web-site.htm 1 root root 9 Jun 21 18:05 index.conf. such as Microsoft FrontPage. Apache searches the DocumentRoot directory for an index.qxd 2/25/05 10:11 AM Page 315 General Configuration Steps 315 GENERAL CONFIGURATION STEPS The configuration file used by Apache is /etc/httpd/conf/httpd.html site1]# ll index. in the httpd. or home. The easiest solution is to create a symbolic link (known as a shortcut to Windows users) called index.htm index. By default.html -> site1]# The l at the very beginning of the the -> indicates the link target. You’ll almost never have to worry about index. Some editors. This isn’t usually a problem if all your HTML files have hyperlinks pointing to files ending in .org in your browser.htm. The problem occurs with Apache not recognizing the topmost index.conf file.html in the /home/www/site1 directory: [root@bigboy [root@bigboy [root@bigboy -rw-rw-r-lrwxrwxrwx index.htm extension.22Harrison_ch20. index. This is needed in Web site managed by the named virtual hosting only. ☞ DocumentRoot: Defines the directory in which the Web pages for the site can be found.htm with index.my-web-site. So for example.html when you enter http://www. As for most Linux applications.htm [root@bigboy tmp]# cd /home/www/site1 site1]# ln -s index. not .htm as FrontPage does.html and Apache again! This example creates a symbolic link to index. if you have a servername of www. or container.htm page. This then enables you to edit or copy the file index. create files with an . you must restart Apache before changes to this configuration file take effect. Apache displays the contents of the file /home/www/site1/index.html being updated automatically.html entry signifies a link and .

You have to make sure all the files and subdirectories in your DocumentRoot have the correct permissions.conf. If you browse your Web site after configuring Apache and get a “403 Forbidden” permissions-related error on your screen. to read the files inside. Create a user with a home directory of /home/www.” has a short script that you can use to recursively set the file permissions in a directory to match those expected by Apache. 2. Recursively change the file ownership permissions of the /home/www directory and all its subdirectories. and Configurations. To do this: 1. The examples in this chapter use the /home/www directory to illustrate how you can place them in other locations successfully. including the Apache’s httpd daemon. This will make all the transferred files automatically have the correct ownership. 3.htm lrwxrwxrwx 1 www users 9 Jun 25 18:05 index.22Harrison_ch20. “Codes. Change the permissions on the /home/www directory to 755.qxd 2/25/05 10:11 AM Page 316 316 The Apache Web Server Chapter 20 The Default File Location By default. The code you need is: [root@bigboy tmp]# useradd -g users www [root@bigboy tmp]# chown -R www:users /home/www [root@bigboy tmp]# chmod 755 /home/www Now test for the new ownership with the ll command: [root@bigboy tmp]# ll /home/www/site1/index.html -> index. Appendix II. It is a good idea to have the files owned by a nonprivileged user so that Web developers can update the files using FTP or SCP without requiring the root password. then your files or directories under your DocumentRoot most likely have incorrect permissions.htm [root@bigboy tmp]# Tip Be sure to FTP or SCP new files to your Web server as this new user. Apache expects to find all its Web page files in the /var/www/html/ directory with a generic DocumentRoot statement at the beginning of httpd. File Permissions and Apache Apache will display Web page files as long as they are world readable. which allows all users. You may also have to use the Directory directive to make Apache .* -rw-rw-r-1 www users 48590 Jun 25 23:43 index. Scripts.

This label is not accessible by Apache. you will get the default Fedora Apache page where your expected Web page should be. When a file is created. The most important types of security label are listed in Table 20.html -rw-r--r-. Same as the httpd_sys_script_ro_t type. Files with this type may be changed by a CGI script in any way.htm extensions. and in some cases. user root creates a directory /home/www/site1 in which the pages for a new Web site will be placed. The context of a file depends on the SELinux label it is given. it inherits the security context of its parent directory. Required for CGI scripts to read files and directories. security contexts become important when Web pages need to be placed in directories that are not the Apache defaults. “Miscellaneous Linux Topics.qxd 2/25/05 10:11 AM Page 317 General Configuration Steps 317 serve the pages once the file permissions have been correctly set.html page created in it.html and . but also allows appending data to files by the CGI script.html [root@bigboy tmp]# ls -Z /home/www/site1/index.” for details. (See Appendix I. Table 20. Using the ls -Z command.html [root@bigboy tmp]# site1 . then this second step becomes unnecessary.) A Web page may have the right permissions. The type required for the execution of CGI scripts. including deletion. then they will inherit the context of that directory and you should have very few problems.1. but the Apache httpd daemon won’t be able to read it unless you assign it the correct security context or daemon access permissions. you can see that the user_home_t security label has been assigned to the directory and the index.22Harrison_ch20. Security Contexts for Web Pages Fedora Core 3 introduced the concept of security contexts as part of the Security Enhanced Linux (SELinux) definition.root root root:object_r:user_home_t /home/www/site1/index. [root@bigboy tmp]# mkdir /home/www/site1 [root@bigboy tmp]# ls -Z /home/www/ drwxr-xr-x root root root:object_r:user_home_t [root@bigboy tmp]# touch /home/www/site1/index. If you decide to place your Web pages in the default /var/www/ directory. Contextrelated configuration errors will give “403 Forbidden” browser messages. In this example.1 SELinux Security Context File Labels Description The type used by regular static Web pages with . HTTP Code httpd_sys_content_t httpd_sys_script_ro_t httpd_sys_script_ra_t httpd_sys_script_rw_t httpd_sys_script_exec_t As expected. If you have your files in the default /home/www directory.

including PERL and PHP. Recognizing the error. By default. [root@bigboy tmp]# tail /var/log/httpd/error_log [Fri Dec 24 17:59:24 2004] [error] [client 216. You can see this when the file /home/www/site1/test.22Harrison_ch20. even though the permissions are correct. CGI scripts are placed in the /var/www/cgi-bin/ directory as defined by the . user root uses chcon with the -R (recursive) and -h (modify symbolic links) qualifiers to modify the label of the directory to httpd_sys_content_t with the -t qualifier.root root root:object_r:httpd_sys_content_t test. [root@bigboy tmp]# chcon -R -h -t httpd_sys_content_t /home/www/site1 [root@bigboy tmp]# ls -Z /home/www/site1/ -rw-r--r-. CGI scripts can be written in a variety of languages.root root root:object_r:httpd_sys_content_t index. Viewing the /var/log/httpd/error_log gives a “Permission Denied” message and the /var/log/ messages file shows kernel audit errors. because new files created in the directory will inherit the SELinux security label of the parent directory. and can be used to do such things as generate new Web page output or update data files.txt [root@bigboy tmp]# ls -Z /home/www/site1/ -rw-r--r-.html file via a Web browser gets a “Forbidden 403” error on your screen.10.txt [root@bigboy tmp]# Security Contexts for CGI Scripts You can use Apache to trigger the execution of programs called Common Gateway Interface (CGI) scripts.119.250] (13)Permission denied: access to / denied [root@bigboy tmp]# tail /var/log/messages Dec 24 17:59:24 bigboy kernel: audit(1103939964. User root won’t have to run the command again for the directory.444:0): avc: denied { getattr } for pid=2188 exe=/usr/sbin/httpd path=/home/www/site1 dev=hda5 ino=73659 scontext=system_u:system_r:httpd_t tcontext=root:object_r:user_home_t tclass=dir [root@bigboy tmp]# SELinux security context labels can be modified using the chcon command. A Web page’s Submit button usually has a CGI script lurking somewhere beneath.qxd 2/25/05 10:11 AM Page 318 318 The Apache Web Server Chapter 20 Accessing the index.txt is created: chcon [root@bigboy tmp]# touch /home/www/site1/test.html -rw-r--r-.root root root:object_r:httpd_sys_content_t index.html [root@bigboy tmp]# Browsing now works without errors.

22Harrison_ch20.cgi” print qq( <html> <head> <meta http-equiv=”Content-Language” content=”en-us”> <meta http-equiv=”Content-Type” content=”text/html”> <title>Linux Home Networking</title> </head> <body> <p align=”center”><font size=”7”>Success!</font></p> </body> </html> ). [root@bigboy tmp]# mkdir -p /home/www/cgi-bin/test [root@bigboy tmp]# chcon -h -t httpd_sys_script_exec_t /home/www/cgi-bin/ [root@bigboy tmp]# mkdir /home/www/cgi-bin/test [root@bigboy tmp]# ls -Z /home/www/cgi-bin drwxr-xr-x root root root:object_r:httpd_sys_script_exec_t test [root@bigboy tmp]# vi /home/www/cgi-bin/test/test. the URL http://192.cgi with the correct executable file permissions. changes the directory’s security context label to httpd_sys_script_exec_t.cgi was created to display the word “Success” on the screen of your Web browser. So.cgi. the PERL script test. which I’ll discuss in ScriptAlias /cgi-bin/ “/var/www/cgi-bin/” In the default case.168.conf file. httpd.qxd 2/25/05 10:11 AM Page 319 General Configuration Steps 319 ScriptAlias directive you’ll find in the more detail later. any URL with the string /cgi-bin/ will trigger Apache to search for an equivalent executable file in this directory.cgi [root@bigboy tmp]# chmod o+x /home/www/cgi-bin/test/test.cgi [root@bigboy tmp]# .1. #!/usr/bin/perl # CGI Script “test. The ScriptAlias directive was set to point to of /var/www/cgi-bin/.100/cgi-bin/test/test. and then creates the script /home/www/cgi-bin/test/test. In the example case. SELinux contexts have to be modified according to the values in Table 20.cgi actually executes the script file /var/www/cgi-bin/test/test. for example.1 for a CGI script to be run in another directory or to access data files. ScriptAlias /cgi-bin/ “/home/www/cgi-bin/” /home/www/cgi-bin/ instead User root creates the /home/www/cgi-bin/ directory.

You use the NameVirtualHost directive in the /etc/httpd/conf/httpd. you can turn them off system wide by editing your /etc/selinux/config file.cgi [root@bigboy tmp]# ls -Z /web/cgi-bin/site2/test.22Harrison_ch20.cgi file needs to be used by a second Web site housed on the same Web server. Named Virtual Hosting You can make your Web server host more than one site per IP address by using Apache’s named virtual hosting feature. ScriptAlias /cgi-bin/ “/web/cgi-bin/site2/” The file inherits the context of its new parent.119. modifying the SELINUX parameter to disabled. but the security context label isn’t copied along with it.qxd 2/25/05 10:11 AM Page 320 320 The Apache Web Server Chapter 20 Accessing the URL http://192. Problems occur when the same test.cgi /web/cgibin/site2/test.conf file to tell Apache which IP addresses will participate in this feature.cgi [root@bigboy tmp]# Permission denied and kernel audit errors occur once more.cgi denied [root@bigboy tmp]# tail /var/log/messages Dec 24 18:36:08 bigboy kernel: audit(1103942168. SELinux will be disabled after your next reboot.100/cgi-bin/test/test.cgi file.cgi dev=hda5 ino=77491 scontext=system_u:system_r:httpd_t tcontext=root:object_r:tmp_t tclass=file [root@bigboy tmp]# Note If you find security contexts too restrictive.549:0): avc: denied { getattr } for pid=2191 exe=/usr/sbin/httpd path=/web/cgibin/site2/test. . you can fix them only by changing the security context of the test.10. [root@bigboy tmp]# tail /var/log/httpd/error_log [Fri Dec 24 18:36:08 2004] [error] [client 216.cgi -rw-r--r-x root root root:object_r:tmp_t /web/cgi-bin/site2/test.cgi is successful. The file is copied to a directory /web/cgi-bin/site2/ governed by the ScriptAlias in the second Web site’s <VirtualHost> container (explained later).168.1.250] (13)Permission denied: access to /cgi-bin/texcelon/test. [root@bigboy tmp]# cp /home/www/cgi-bin/test/test.

and ServerName when making a decision as to which content to send to the remote user’s Web browser. You can also list secondary domain names that will serve the same content as the primary ServerName using the ServerAlias directive. NameVirtualHost 97. The DocumentRoot directive defines the directory that contains the index page for that site.253. not site #1 or site #2) </VirtualHost> <VirtualHost 97.253.org Directives for site #1 </VirtualHost> <VirtualHost 97. In this case.253. A <VirtualHost> with a specific IP address always gets higher priority than a <VirtualHost> statement with an * intended to cover the same IP address.26 <VirtualHost *> Default Directives.26> servername www. You can also have multiple NameVirtualHost directives.158. then Apache uses the first <VirtualHost> in the list that matches the target IP address of the request.26.158. IP-Based Virtual Hosting The other virtual hosting option is to have one IP address per Web site. You must specify the IP address for which each <VirtualHost> container applies.my-web-site.org Directives for site #2 </VirtualHost> Be careful with using the asterisk in other containers. If there is no match. in cases where your Web server has more than one IP address. In the code that follows.253. try to limit the use of your <VirtualHost *> statements to the beginning of the list to cover any other IP addresses your server may have. To get consistent results.158. each with a single IP address.26> servername www. which is also known as IP-based virtual hosting.qxd 2/25/05 10:11 AM Page 321 General Configuration Steps 321 The <VirtualHost> containers in the file then tell Apache where it should look for the Web pages used on each Web site. even if the ServerName directive doesn’t match.another-web-site. Apache searches for a perfect match of NameVirtualHost. Named Virtual Hosting Example Consider an example in which the server is configured to provide content on 97.22Harrison_ch20. This is why the first <VirtualHost> statement contains an asterisk: to indicate it should be used for all other Web queries. notice that within each <VirtualHost> container you specify the primary Web site domain name for that IP address with the ServerName directive. <VirtualHost>. you will not have a .158. (In other words.

qxd 2/25/05 10:11 AM Page 322 322 The Apache Web Server Chapter 20 NameVirtualHost directive for the IP address.253. Web surfers get the site1 content if they try to access the Web server on any of its other IP addresses. <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> IP Virtual Hosting Example: Wild Card and IP Addresses In this example. the ServerName directive isn’t needed in each <VirtualHost> container.253. IP Virtual Hosting Example: Single Wild Card In this example.158. Apache interprets it as an overlap of name-based and IP-based <VirtualHost> .253. and you must only have a single Also. then wild cards won’t work. The Apache SSL module demands at least one explicit <VirtualHost> directive for IP-based virtual hosting. <VirtualHost> container per IP address.22Harrison_ch20.27. system administrators commonly replace the IP address in the <VirtualHost> and NameVirtualHost directives with the * wildcard character to indicate all IP addresses. unlike in named virtual hosting. Apache listens on all interfaces. but gives different content for addresses 97. which is used frequently in credit card and shopping cart Web pages.158. because there is only one Web site per IP address. but gives the same content.26> DocumentRoot /home/www/site2 </VirtualHost> <VirtualHost 97. <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> <VirtualHost 97. When you use wild cards. If you installed Apache with support for secure HTTPS/SSL. Apache displays the content in the first <VirtualHost *> directive even if you add another right after it.253.26 and 97. Apache listens on all interfaces.158.158.27> DocumentRoot /home/www/site3 </VirtualHost> Virtual Hosting and SSL Because it makes configuration easier. Apache also seems to enforce the single <VirtualHost> container per IP address requirement by ignoring any ServerName directives you may use inside it.

I’ll walk you through an example scenario.158.conf file.org.org. The administrator wants to be able to get to test.158.22Harrison_ch20. Don’t use virtual hosting statements with wild cards except for the very first <VirtualHost> directive that defines the Web pages to be displayed when matches to the other <VirtualHost> directives cannot be found.my-web-site.my-web-site. The domain www.my-web-business.158.my-web-site.qxd 2/25/05 10:11 AM Page 323 Configuration: Multiple Sites and IP Addresses 323 directives and gives error messages because it can’t make up its mind about which method to use: Starting httpd: [Sat Oct 12 21:21:49 2002] [error] VirtualHost _default_:443 -.28> Directives for site that also run on SSL </VirtualHost> CONFIGURATION: MULTIPLE SITES AND IP ADDRESSES To help you better understand the edits needed to configure the /etc/httpd/conf/httpd. The parameters are: ☞ The Web site’s systems administrator previously created DNS entries for www. you’ll also see the error: Bad request! Your browser (or proxy) sent a request that this server could not understand.253.org on all the IP addresses.another-web-site. Here is an example: NameVirtualHost * <VirtualHost *> Directives for other sites </VirtualHost> <VirtualHost 97.26 on this Web server.org.253.mixing * ports and non-* ports with a NameVirtualHost address is not supported. . and default.253. please contact the webmaster The best solution to this problem is to use wild cards more sparingly.org is also configured to point to alias IP address 97. If you think this is a server error. www. proceeding with undefined results If you try to load any Web page on your Web server.org to map to the IP address 97. my-web-site.27.

org falls in this category.org my-web-site.26 is therefore required. Table 20. ☞ Traffic going to www. ☞ Named virtual hosting will be required for 97. ☞ Traffic to test.253.org ServerAlias my-web-site.158.253.158.my-website.158. A NameVirtualHost directive for 97. my-web-site.253.26 NameVirtualHost 97.158.26 as in this case we have a single IP address serving different content for a variety of domains.158.my-web-site.my-web-site.253.26> DocumentRoot /home/www/site2 ServerName www.158.27 97.my-web-site.27 # # Match a webpage directory with each website # <VirtualHost *> DocumentRoot /home/www/site1 </VirtualHost> <VirtualHost 97.org.conf file: ServerName localhost NameVirtualHost 97.158.26 Site3 Site1 Site1 Name based (Wild card) Name based Name based All other domains How do these requirements translate into code? Here is a sample snippet of a working httpd.org Web Hosting Scenario Summary IP address 97.27 97.my-web-business.2 summarizes these requirements.org www.158. Table 20. www. Hitting these URLs causes Apache to display the contents of index.another-web-site. ☞ All other domains pointing to this server that don’t have a matching ServerName directive will get Web pages from the directory defined in the very first <VirtualHost> container: directory site1.org must get content from subdirectory site3.html in this directory.org.my-web-business.org </VirtualHost> .my-web-site.org test.another-web-site.qxd 2/25/05 10:11 AM Page 324 324 The Apache Web Server Chapter 20 ☞ Traffic to www. and www.my-web-site.org.253.253.26 Directory Site2 Type of Virtual Hosting Name based 97.org will get content from directory site4.253. Site default.158.org default.253.my-web-business.org must get content from subdirectory site2.org www.2 Domain www.22Harrison_ch20.my-web-site.253.

deny Allow from all </Limit> <LimitExcept GET POST OPTIONS> Order deny.27> DocumentRoot /home/www/site3 ServerName test. Testing Your Web Site Before DNS Is Fixed You may not be able to wait for DNS to be configured correctly before starting your project. The easiest way to temporarily bypass this is to modify the hosts file on the Web developer’s client PC or workstation (not the Apache server).org.22Harrison_ch20.my-web-site.qxd 2/25/05 10:11 AM Page 325 Configuration: Multiple Sites and IP Addresses 325 <VirtualHost 97.another-web-site.253. # <Directory “/home/www/*”> Order allow.org and another-web-site. The last section of this configuration snippet has some additional statements to ensure read-only access to your Web pages with the exception of Web-based forms using POSTs (pages with “submit” buttons). such as my-web-site.allow Deny from all </LimitExcept> </Directory> These statements would normally be found at the very bottom of the file where the virtual hosting statements reside.deny Allow from all AllowOverride FileInfo AuthConfig Limit Options MultiViews -Indexes SymLinksIfOwnerMatch IncludesNoExec <Limit GET POST OPTIONS> Order allow.253. Chapter 18 shows you how to configure multiple domains.org </VirtualHost> # # Make sure the directories specified above # have restricted access to read-only.conf file for the changes to take effect on the running process.org </VirtualHost> <VirtualHost 97. on your DNS server.158. Remember to restart Apache every time you update the httpd.27> DocumentRoot /home/www/site4 ServerName www. Note You will have to configure your DNS server to point to the correct IP address used for each of the Web sites you host.158. .

my-web-site. you create a subdirectory named /home/www/site1/example under www. Apache will default to giving a listing of all the files in that subdirectory. Now you’ll be able to view the contents of the file my-example. You can disable the directory listing by using a Indexes option in the <Directory> directive for the DocumentRoot like this: <Directory “/home/www/*”> .org’s DocumentRoot of /home/www/site1/..158.my-web-site.html pages in each subdirectory under your DocumentRoot directory. Users attempting to access the nonexistent index page will now get a “403 Access Denied” message. The Windows equivalent of the Linux /etc/hosts file is named C:\ WINDOWS\system32\drivers\etc\hosts. for example.253.org is listed in the file. PCs and Linux workstations query the hosts file first before checking DNS.org/example/my-example..org/example www. This allows Web surfers to double-check the locations of files through their browsers.html in this subdirectory if you point your browser to: http://www.26 www. Note When setting up a yum server it’s best to enable directory listings for the RPM subdirectories..my-web-site. so if a value for www. .22Harrison_ch20.my-web- Apache lists all the contents of the files in the example directory if it can’t find the index. .org Do not remove the localhost entry in this file.html site. . Say.html file.org/example. such as Notepad.. Here you could add an entry similar to: 97...my-web-site.my-web-site. Options MultiViews -Indexes SymLinksIfOwnerMatch IncludesNoExec Remember to restart Apache after the changes. You need to open and edit it with a text editor. Disabling Directory Listings Be careful to include any index.qxd 2/25/05 10:11 AM Page 326 326 The Apache Web Server Chapter 20 By default. that’s what the client will use. because if one isn’t found. If curious surfers decide to see what the index page is for they would type the link: http://www.

You can place this statement in the httpd.. transparently uncompressing the data and presenting it on the screen.115 . .htm instead of a generic “404 File Not Found” message: ErrorDocument 404 /missing.. YComp 5.conf file.qxd 2/25/05 10:11 AM Page 327 Using Data Compression on Web Pages 327 Handling Missing Pages You can tell Apache to display a predefined HTML file whenever a surfer attempts to access a non-index page that doesn’t exist. [root@ bigboy tmp]# . You can see the missing..119.. First you need to load Apache version 2’s deflate module in your httpd..25.conf file and then use Location directives to specify which type of files to compress.linuxhomenetworking.linuxhomenetworking. USING DATA COMPRESSION ON WEB PAGES Apache also has the ability to dynamically compress static Web pages into gzip format and then send the result to the remote Web surfers’ Web browser.0. you will be able to verify from your /var/log/httpd/access_log file that the sizes of the transmitted HTML pages have shrunk.com/bogus-file. 67..5. which will make Apache display the contents of missing..2.htm. MSIE 5. This can significantly reduce bandwidth charges if you are paying for Internet access by the megabyte. Compare the file sizes in this Apache log: [root@ bigboy tmp]# grep dns-static /var/log/httpd/access_log . Windows NT 4. After making these modifications and restarting Apache..com/missing.0..linuxhomenetworking.6)” .[15/Feb/2003:23:06:51 -0800] “GET /dns-static.htm HTTP/1. Most current Web browsers support this format.htm file I use by trying the nonexistent link: http://www.0.0 (compatible. .htm Notice that this gives the same output as: http://www.22Harrison_ch20.htm” “Mozilla/4.1” 200 15190 “http://www. AT&T CSM6.htm Remember to put a file with this name in each DocumentRoot directory.com/sendmail.

78. Note Fedora’s version of httpd.so <Location /> # Insert filter SetOutputFilter DEFLATE # Netscape 4.190 bytes—that’s almost 80% compression.qxd 2/25/05 10:11 AM Page 328 328 The Apache Web Server Chapter 20 and the corresponding directory listing: [root@ bigboy tmp]# ll /web-dir/dns-static. Remember to restart Apache when you do.. but it is fine BrowserMatch \bMSIE !no-gzip !gzip-only-text/html # Don’t compress images SetEnvIfNoCase Request_URI \ \.(?:gif|jpe?g|png)$ no-gzip dont-vary # Make sure proxies don’t deliver the wrong content Header append Vary User-Agent env=!dont-vary </Location> . Compression Configuration Example You can insert these statements just before your virtual hosting section of your httpd. however. This means that the LoadModule line (the first line of the example snippet) is not required for Fedora.htm -rw-r--r-1 user group 78350 Feb 15 00:53 /home/www/ccie/dns-static.06-4.conf loads the compression module mod_deflate by default.08 have some more problems BrowserMatch ^Mozilla/4\..htm [root@bigboy tmp]# As you can see. BrowserMatch ^Mozilla/4 gzip-only-text/html # Netscape 4.22Harrison_ch20.0[678] no-gzip # MSIE masquerades as Netscape.conf file to activate the compression of static pages. The location statements are required.x has some problems. LoadModule deflate_module modules/mod_deflate.350 bytes shrunk to 15.

firewalls frequently don’t allow access from their protected network to IP addresses that they masquerade on the outside. Because of NAT (network address translation).158.1.100 NameVirtualHost 97.158.X network. and also when accessing the specific IP address 192.26 when visiting www.253.26 via NAT/masquerading.mysite. .168. www. Users on the Internet need to access IP address 97. There is a two-part solution to this problem: Step 1: Configure Virtual Hosting on Multiple IPs You can configure Apache to serve the correct content when accessing www.another-web-site.org and users on your home network need to access IP address 192.org.253.168.com or www.1.mysite. 192.org from the outside. Linux Web server Bigboy has an internal IP address of 192. then you may encounter problems when trying to access www. Chapter 18 explains how to do this in detail. If you are on the inside.another-web-site.253.1.another-web-site.168. You can configure your DNS server to use views which makes your DNS server give different results depending on the source IP address of the Web surfer’s PC doing the query.com of www.158.1.100 when visiting the same site.org ServerAlias bigboy. you may find it impossible to hit URLs that resolve in DNS to 97. but the firewall presents it to the world with an external IP address of 97.158.qxd 2/25/05 10:11 AM Page 329 Apache Running on a Server Behind a NAT Firewall 329 APACHE RUNNING ON A SERVER BEHIND A NAT FIREWALL If your Web server is behind a NAT firewall and you are logged on a machine behind the firewall as well.100.1.168.org </VirtualHost> Step 2: Configure DNS Views You now need to fix the DNS problem that NAT creates.100 from the inside. Fortunately Apache allows you to specify multiple IP addresses in the <VirtualHost> statements to help you overcome this problem.26.1.158.my-web-site.168.168.22Harrison_ch20.253. For example.my-web-site.100 97. Here is an example: NameVirtualHost 192.253.26> DocumentRoot /www/server1 ServerName www.26 <VirtualHost 192.

The require user statement tells Apache that only user peter in the . switch on the command line. then you can edit your PC’s hosts file as a quick and dirty temporary solution to the problem. Remember that this will fix the problem on your PC alone. I recommend placing the file in your /etc/httpd/conf directory. You have to specify the location of the password file. Here is an example for a first user named peter and a second named paul: [root@bigboy tmp]# htpasswd -c /etc/httpd/conf/.htpasswd AuthGroupFile /dev/null AuthName EnterPassword AuthType Basic require user peter Remember this password protects the directory and all its subdirectories. Make the .htpasswd file readable by all users: [root@bigboy tmp]# chmod 644 /etc/httpd/conf/.qxd 2/25/05 10:11 AM Page 330 330 The Apache Web Server Chapter 20 Tip If you have to rely on someone else to do the DNS change. Create a .htpasswd 3. you have to include a -c.htpasswd file. Use Apache’s htpasswd password utility to create username/password combinations independent of your system login password for Web page access.htpasswd file should . and if it doesn’t yet exist. The AuthUserFile tells Apache to use the . I know people who allow normal access to their regular Web pages.22Harrison_ch20. away from the DocumentRoot tree where Web users could possibly view it. but require passwords for directories or pages that show MRTG or Webalizer data. This example shows how to password protect the /home/www directory: 1.htpasswd paul New password: Re-type new password: Adding password for user paul [root@bigboy tmp]# 2. PROTECTING WEB PAGE DIRECTORIES WITH PASSWORDS You can password protect content in both the main and subdirectories of your DocumentRoot fairly easily.htpasswd peter New password: Re-type new password: Adding password for user peter [root@bigboy tmp]# [root@bigboy tmp]# htpasswd /etc/httpd/conf/.htaccess file in the directory to which you want password control with these entries: AuthUserFile /etc/httpd/conf/. or create.

AuthType Basic instructs Apache to accept basic unencrypted passwords from the remote users’ Web browser.d directory are read and automatically appended to the configuration in the httpd. then place the NameVirtualHost statements in the corresponding /etc/httpd/conf.conf file.253.D DIRECTORY Files in the /etc/httpd/conf.conf file has an AllowOverride statement in a <Directory> directive for any directory in the tree above /home/www. and you don’t have to refer to them in the httpd.htaccess 5. it’ll need to remain in the main httpd.d directory file. Back up your httpd. Now.26 DocumentRoot /home/www </VirtualHost> 7.htaccess file in the directory /home/www: [root@bigboy tmp]# chmod 644 /home/www/. If it is shared. in which a Web server has to host many Web sites. all directories below /var/www/ require password authorization: <Directory /home/www/*> AllowOverride AuthConfig </Directory> 6. when you try accessing the Web site. Restart Apache.qxd 2/25/05 10:11 AM Page 331 The /etc/httpd/conf. Restart Apache. you can create one configuration file per Web site each with its own set of <VirtualHost> and <Directory> containers. and test.htpasswd users to have access. Remove the corresponding directives from the httpd. Set the correct file protections on your new . replace this line with require valid-user. To do this correctly: 1. The files located in the /etc/httpd/conf. Make sure your /etc/httpd/conf/http. 3. Create the files located in this directory that contain the Apache required <VirtualHost> and <Directory> containers and directives. 4.d Directory 331 have access. . Make sure that you have a <VirtualHost> directive that defines access to /home/www or another directory higher up in the tree: <VirtualHost *> ServerName 97. you’ll be prompted for a password. If each site has a dedicated IP address.d directory don’t have to have any special names. 2. THE /ETC/HTTPD/CONF. in case you make a mistake. This can make Web site management much simpler. In this example below.22Harrison_ch20.conf file every time Apache is restarted.conf file. If you want all . 4.conf file.158.conf file. In complicated configurations. 5.

“Simple Network Troubleshooting. . Please refer to the “General Configuration Steps” section for further details.22Harrison_ch20. or port forwarding rules to your Web server. Other sources of failure include Apache not being started at all. not the Web server software itself. It is best to test this from both inside your network and from the Internet. Troubleshooting with TELNET is covered in Chapter 4.html file in your Web site’s DocumentRoot directory. or networkrelated failures. NAT. A sure sign of problems related to security context are “avc: denied” messages in your /var/log/messages log file. Most of the errors you’ll encounter will probably be related to incompatible syntax in the <VirtualHosts> statement caused by typing errors. The first is the lack of an index. Lack of connectivity could be caused by a firewall with incorrect permit. Testing Basic HTTP Connectivity The very first step is to determine whether your Web server is accessible on TCP port 80 (HTTP). If you can connect on port 80 but no pages are being served. then the problem is usually due to a bad Web application. you’ll find errors in the /var/log/httpd/error_log file during normal operation or displayed on the screen when Apache starts up.qxd 2/25/05 10:11 AM Page 332 332 The Apache Web Server Chapter 20 TROUBLESHOOTING APACHE Troubleshooting a basic Apache configuration is fairly straightforward.” Browser 403 Forbidden Messages Browser 403 Forbidden messages are usually caused by file permissions and security context issues. Nov 21 20:41:23 bigboy kernel: audit(1101098483.html dev=hda5 ino=12 scontext=root:system_r:httpd_t tcontext=root:object_r:home_root_t tclass=file Only the Default Apache Page Appears When only the default Apache page appears. The second cause is usually related to an incorrect security context for the Web page’s file. the server being down.897:0): avc: denied { getattr } for pid=1377 exe=/usr/sbin/httpd path=/home/www/index. Please refer to the “General Configuration Steps” section for further details. there are two main causes.

For the simple virtual hosting example above.name:80 ServerName localhost The Apache Status Log Files The /var/log/httpd/access_log file is updated after every HTTP query and is a good source of general purpose information about your Web site.conf.rpmnew file. There is a fixed formatting style with each entry separated by spaces or quotation marks. 2. The new version 2. Table 20.conf.conf-version-1.conf file that references localhost instead of the default new. or else you’ll get an error similar to these when starting httpd: Starting httpd: httpd: Could not determine the server’s fully qualified domain name.3 lists the layout. NameVirtualHost. and VirtualHost containers from the old file and place them in the new httpd. Server Name Errors All ServerName directives must list a domain that is resolvable in DNS.conf file with another name. Copy the httpd.0.0. Copy the ServerName.1.host.x for example.conf.qxd 2/25/05 10:11 AM Page 333 Troubleshooting Apache 333 Incompatible /etc/httpd/conf/http. Restart Apache.name:80: #ServerName new. 3. it would be easiest to: 1.rpmnew.X. and name it httpd.X default configuration file is stored in /etc/httpd/conf/httpd.conf Files When Upgrading Your old configuration files will be incompatible when upgrading from Apache version 1.1 for ServerName Starting httpd: [Wed Feb 04 21:18:16 2004] [error] (EAI 2)Name or service not known: Failed to resolve server name for 192.host. httpd.16.or specify an explicit ServerName You can avoid this by adding a default generic ServerName directive at the top of the httpd.conf.100 (check DNS) -. Save the old httpd. using 127.22Harrison_ch20. .rpmnew file.3 to Apache 2.

Table 20.com/sendmail.22Harrison_ch20.. Successful request. 67.htm returning a successful 200 status code.6)” The HTTP status code can provide some insight into the types of operations surfers are trying to attempt and may help to isolate problems with your pages.[15/Feb/2003:23:06:51 -0800] “GET /dns-static.119.5.115 on February 15 looked at the Web page /dns-static.0. Table 20. but the Web page requested hasn’t been modified since the current version in the remote Web browser’s cache. Frequently occurs when a surfer is browsing back and forth on a site. even though it has universal read access. YComp 5.htm using Microsoft Internet Explorer version 5. This means the Web page will not be sent to the remote browser. The amount of data sent was 15190 bytes and the surfer got to the site by clicking on the link http://www. linuxhomenetworking. 404 errors are generated when someone tries to access a Web page that doesn’t exist anymore. Someone entered an incorrect username or password on a password-protected page.htm HTTP/1.115 . AT&T CSM6.3 Apache Log File Format Description IP address of the remote Web surfer Time stamp HTTP query including the Web page served HTTP result code The amount of data in bytes sent to the remote Web browser The Web page that contained the link to the page served The version of the Web browser used to get the page Separator Spaces Square Brackets [] Quotes “” Spaces Spaces Quotes “” Quotes “” Field Number 1 2 3 4 5 6 7 Upon examining the entry. not the operation of the Apache server. This could be caused by incorrect URL links in other pages on your site. HTTP Code 200 304 401 403 .0.htm” “Mozilla/4.5.com/sendmail. File permissions prevent Apache from reading the file.25. Unauthorized access.119. it will just use its cached version instead. Windows NT 4. For example. Often occurs when the Web page file is owned by user root. MSIE 5.25. you can determine that someone at IP address 67.0.2.4 has some of the more common examples.qxd 2/25/05 10:11 AM Page 334 334 The Apache Web Server Chapter 20 Table 20.linuxhomenetworking.4 HTTP Status Codes Description Successful request. Forbidden.1” 200 15190 “http://www.0 (compatible.

qxd 2/25/05 10:11 AM Page 335 Conclusion 335 HTTP Code 404 500 Description Not found. “Configuring Linux Mail Servers. there is no standardized formatting. a valuable quality when replies need to be complex. Page requested doesn’t exist. Unlike the /var/log/httpd/access_log file. which can be very useful. Check your error_log file for further details on the script’s error message. Typical errors that you’ll find here are HTTP queries for files that don’t exist or forbidden requests for directory listings. has the advantage of being able to relay documents and other information without interrupting the addressee. Frequently generated by CGI scripts that fail from bad syntax. E-mail. even the best Web sites can be impersonal as they frequently only provide information that the designer expects the visitor to need. the /var/log/httpd/error_log file most likely lists the cause of the problem. . such as IP telephony and instant messaging. The /var/log/httpd/error_log file also is the location where CGI script errors are written.22Harrison_ch20. Internal server error. although ancient in comparison to newer personalized interactive Internet technologies. CONCLUSION Web sites both personal and commercial can be very rewarding exercises as they share your interests with the world and allow you to meet new people with whom to develop friendships or transact business. Many times CGI scripts fail with a blank screen on your browser. Unfortunately. No Web site should be without one. The Apache Error Log Files The /var/log/httpd/error_log file is a good source for error information. The file will also include Apache startup errors. Chapter 21. This allows them to schedule a response when they are better prepared to answer.” explains how to configure a Linux e-mail server to reduce spam and provide personalized addresses across multiple domains.

22Harrison_ch20.qxd 2/25/05 10:11 AM Page 336 .

In a home environment. then a dedicated mail server will probably be required. are also gaining popularity with Linux. sendmail can handle both incoming and outgoing mail for your domain. such as postfix and qmail. You’ll also learn how to retrieve and send mail via your mail server using a mail client such as Outlook Express or Evolution. This chapter will show you how to use sendmail to create a mail server that will relay your mail to a remote user’s mailbox or incoming mail to a local mailbox.org) is to use the MX record in the configuration zone file to state the hostname of the server that will handle the mail for the domain.23Harrison_ch21.qxd 2/25/05 10:38 AM Page 337 C H A P T E R 21 Configuring Linux Mail Servers IN THIS CHAPTER ☞ ☞ ☞ ☞ Configuring sendmail Fighting Spam Configuring Your POP Mail Server Conclusion E-mail is an important part of any Web site you create. How sendmail Works As stated before. The most popular Unix mail transport agent is sendmail. Take a closer look. a free Web-based e-mail service may be sufficient. but if you are running a business. The steps used to convert a Linux box into a sendmail mail server will be explained here. CONFIGURING SENDMAIL One of the tasks in setting up DNS for your domain (my-web-site. 337 . but others.

Linux workstation users can use the same programs to access their mail remotely. such as mail. sendmail first uses DNS to get the MX record for the other domain. then your mail server could be commandeered to relay spam. and it contains more understandable instructions for systems administrators to use. If the mail is being sent to another domain.23Harrison_ch21.mc. Users retrieve their mail from the mail server using client software.cf file. sendmail inspects this header and then acts accordingly. It then attempts to relay the mail to the appropriate destination mail server using the Simple Mail Transport Protocol (SMTP). sendmail Macros When mail passes through a sendmail server the mail routing information in its header is analyzed. Simple sendmail security will be covered later. then sendmail places the message in that person’s mailbox so that they can retrieve it using one of the methods above. Mail isn’t actually sent directly to the user’s PC. according to the desires of the systems administrator. Using a series of highly complicated regular expressions listed in the /etc/mail/sendmail. you must always run the macros and restart sendmail for the changes to take effect. the PC of user A can delegate the SMTP processing to the mail server. such as Evolution.com domain. One of the main advantages of mail relaying is that when a PC user A sends mail to user B on the Internet. and sometimes modified.mc was created. PC and Linux workstation users configure their e-mail software to make the mail server their outbound SMTP mail server. After editing sendmail. These are then interpreted by a number of macro routines to create the sendmail. such as Microsoft’s Outlook or Outlook Express. or a GUI client.org) eventually arrives at your mail server and sendmail then processes it and deposits it in the mailbox file of the user’s Linux account. . Linux users logged into the mail server can read their mail directly using a text-based client. that supports either the POP or IMAP mail retrieval protocols. Mail sent to each of these users (username@my-web-site.cf file. Outgoing Mail The process is different when sending mail via the mail server. a much simpler file named /etc/sendmail. Note If mail relaying is not configured properly.qxd 2/25/05 10:38 AM Page 338 338 Configuring Linux Mail Servers Chapter 21 Incoming Mail Usually each user in your home has a regular Linux account on your mail server. If the mail is destined for a local user in the mysite.cf file. In recognition of the complexity of the /etc/mail/sendmail.

[root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make make If there have been no changes to the files in /etc/mail since the last time was run. to the sendmail. then you’ll get an error like this: [root@bigboy mail]# make make: Nothing to be done for `all’. The m4 usage is simple. Notice that sendmail is sensitive to the quotation marks used in the m4 macro directives. A typical example is: FEATURE(`virtusertable’.23Harrison_ch21.cf file.`hash -o /etc/mail/virtusertable. [root@bigboy mail]# The make command actually generates the sendmail. which would normally go to the screen.db’)dnl The keywords usually define a subdirectory of /usr/share/sendmail-cf. the macro name is /usr/share/ sendmail-cf/feature/virtusertable. Once you finish editing the sendmail. and redirect the output.mc file. followed by a subdirective and in some cases arguments.cf I’ll discuss many of the features of the sendmail.mc > /etc/mail/sendmail. you just specify the name of the macro file as the argument.m4.cf file using the m4 command. They open with a grave mark and end with a single quote: FEATURE(`masquerade_envelope’)dnl Some keywords. such as define for the definition of certain sendmail variables and MASQUERADE_DOMAIN. have no corresponding directories with matching macro files. such as DOMAIN. in this case sendmail.mc file later in the chapter.mc. . in which the macro may be found and the subdirective is usually the name of the macro file itself.qxd 2/25/05 10:38 AM Page 339 Configuring sendmail 339 Each sendmail. you can then execute the make command while in the /etc/mail directory to regenerate the new sendmail. FEATURE.mc directive starts with a keyword. or OSTYPE.db’ is being passed to it. So in the example. # m4 /etc/mail/sendmail.cf file with the > redirector symbol. The macros in the /usr/share/sendmail-cf/m4 directory deal with these. and the instruction `\ hash -o /etc/ mail/virtusertable.

10-1. “Installing RPM Software. and m4 software RPMs are installed. you’ll see that sendmail uses a variety of configuration files that require different treatments for their commands to take effect. [root@bigboy tmp]# pgrep sendmail You should get a response of plain old process ID numbers. remember that the filename usually starts with the software package name by a version number. and restart sendmail after booting.rpm. stop. the .d/sendmail restart It first runs the make command. You will need to make sure that the sendmail.1.23Harrison_ch21. This little script encapsulates all the required post configuration steps: #!/bin/bash cd /etc/mail make newaliases /etc/init. Starting sendmail You can use the chkconfig command to get sendmail configured to start at boot: [root@bigboy tmp]# chkconfig sendmail on To start.) When searching for the RPMs. as in sendmail8. sendmail-cf. You can also test whether the sendmail process is running with the pgrep command.1.i386.12. use: [root@bigboy tmp]# service sendmail start [root@bigboy tmp]# service sendmail stop [root@bigboy tmp]# service sendmail restart Remember to restart the sendmail process every time you make a change to the configuration files for the changes to take effect on the running process.cf file from sendmail.qxd 2/25/05 10:38 AM Page 340 340 Configuring Linux Mail Servers Chapter 21 Installing sendmail Most Red Hat and Fedora Linux software products are available in the RPM format.mc file and compiles supporting configuration files in the /etc/mail directory according to the instructions in the file /etc/mail/Makefile. (Chapter 6. How to Restart sendmail After Editing Your Configuration Files In this chapter. which creates a new sendmail.” will tell you how.

(this will be covered later).cf file. Note the file sendmail.mc file.1. Use this command to make the script executable: chmod 700 filename You’ll need to run the script each time you change any of the sendmail configuration files. in most cases you won’t have to edit this file very often. it reads sendmail.0. How to Put Comments in sendmail.mc configuration file: These statements are disabled by dnl commenting: dnl DAEMON_OPTIONS(`Port=smtp.0 and higher and Fedora Core.0.cf file is used for versions up to 7. you can define most of sendmail’s configuration parameters in the /etc/mail/sendmail. but it is still often viewed as an intimidating task with its series of structured directive statements that get the job done.1. The sendmail.mc File As discussed earlier.Addr=127.mc file doesn’t use this character for commenting.mc In most Linux configuration files a # symbol is used at the beginning of a line to convert it into a comment line or to deactivate any commands that may reside on that line. The /etc/sendmail. but instead uses the string “dnl”. Name=MTA’) . Configuration of the sendmail.Addr=127.mc file. configuration file and really is much easier to fool around with without getting burned. Fortunately.cf. which is then used by the m4 macros to create the /etc/mail/sendmail.cf for its configuration.cf file is located in different directories depending on the version of Red Hat you use. and /etc/mail/sendmail. Here are some valid examples of comments used with the sendmail.0.23Harrison_ch21. Name=MTA’) dnl # DAEMON_OPTIONS(`Port=smtp.3. When sendmail starts. This may not be a good idea in a production system.0. The line in the script that restarts sendmail is only needed if you have made changes to the /etc/mail/sendmail. and then restarts sendmail. The sendmail. but I included it so that you don’t forget.mc is a more user friendly The /etc/mail/sendmail.cf is used for versions 8.qxd 2/25/05 10:38 AM Page 341 Configuring sendmail 341 It then generates new e-mail aliases with the newaliases command.mc file is much simpler than configuration of sendmail.

1.org. Finally.qxd 2/25/05 10:38 AM Page 342 342 Configuring Linux Mail Servers Chapter 21 This statement is incorrectly disabled: # DAEMON_OPTIONS(`Port=smtp.conf File The sendmail program expects DNS to be configured correctly on the DNS server. The program also expects the files used by the mail server’s DNS client to be configured correctly.” or Chapter 19.org and the IP address of the DNS server is 192.my-web-site.my-site.1. Name=MTA’) Configuring DNS for sendmail Remember that you will never receive mail unless you have configured DNS for your domain to make your new Linux box mail server the target of the DNS domain’s MX record.com bigboy.168.0.Addr=127.168. For example.org nameserver 192. there must be directives that look like this: domain my-web-site.conf file in which there must be a domain directive that matches one of the domains the mail server is expected to handle mail for. The first one is the /etc/resolv.0.org must correctly resolve to the IP address of one of the mail server’s interfaces.my-website. “Dynamic DNS. Configure the /etc/resolv. The MX record for your domain must point to the IP address of the mail server. then bigboy. Configure Your Mail Server’s Name In DNS You first need to make sure that your mail server’s name resolves in DNS correctly. if the mail server is handling mail for my-web-site.org has address 192.Addr=127. “Configuring DNS.0.100 .1.1. Name=MTA’) This statement is active: DAEMON_OPTIONS(`Port=smtp. sendmail expects a nameserver directive that points to the IP address of the DNS server the mail server should use to get its DNS information.1. You can test this using the host command: [root@smallfry tmp]# host bigboy.” for details on how to do this. See either Chapter 18. For example.23Harrison_ch21.100.168.0. if your mail server’s name is Bigboy and you intend for it to mostly handle mail for the domain my-web-site.100 [root@smallfry tmp]# You will need to fix your DNS server’s entries if the resolution isn’t correct.

then the /etc/hosts file will also have to be correctly configured.org localhost.localdomain. As mail must be sent to a target IP address even when there is no NIC in the box.org. Linux does not function properly if the 127.qxd 2/25/05 10:38 AM Page 343 Configuring sendmail 343 An incorrectly configured resolv.localdomain localhost bigboy The entry for 127. you can either create one or use the one offered by your ISP. To do this. you can add any other aliases your host may have to the end of the line. the mail server for the my-web-site.my-web-site.localdomain. you need to process the sendmail.1 entry in /etc/hosts doesn’t also include localhost and localhost.my-web-site. In the case above. Converting From a Mail Client to a Mail Server All Linux systems have a virtual loopback interface that lives only in memory with an IP address of 127.0.my-web-site.1 must always be followed by the fully qualified domain name (FQDN) of the server. Finally.org domain. How to Configure Linux sendmail Clients All Linux mail clients in your home or company need to know which server is the mail server. Once this is done.1 bigboy.0. Then you must have an entry for localhost and localhost. If the sendmail server is a Linux server. the mail server has been set to mail.0. This is configured in the sendmail. define(`SMART_HOST’. Here is a brief example of the first line you should expect to see in it: 127. run the restarting script from earlier in the chapter. sendmail therefore uses the loopback address to send mail between users on the same Linux server.`mail.conf file can lead to errors when running the m4 command to process the information in your sendmail.0.mc file by setting the SMART_HOST statement to include the mail server.org’) If you don’t have a mail server on your network. it would be bigboy. To .0.1.mc file and restart sendmail.0.mc file: WARNING: local host name (smallfry) is not qualified.myweb-site.0.23Harrison_ch21. fix $j in config file The /etc/hosts File The /etc/hosts file also is used by DNS clients and also needs to be correctly configured. In the example below.org.0.

0. You can see the interfaces on which sendmail is listening with the netstat command.0.0.0.0.0:* LISTEN [root@bigboy tmp]# 2. If sendmail is listening on the loopback interface only.0. you should comment out the daemon_options line in the /etc/mail/sendmail. Chapter 18 shows how to create your own internal domain just for this purpose.0. Edit sendmail. It is also good practice to take precautions against spam by not accepting mail from domains that don’t exist by commenting out the accept_ unresolvable_domains feature too.1 (loopback): [root@bigboy tmp]# netstat -an | grep :25 | grep tcp tcp 0 0 127. as well: 1. the laptop and users on computers that do not have 24x7 DNS do need this. In the sample network. Note If your server has multiple NICs and you want it to listen to only one of them. sendmail needs to be configured to listen for messages on NIC interfaces.0. Comment this out if you want to accept email over the network.1 and not on any other network devices.mc file with dnl statements.0. you use netstat and grep for 25 to see a default configuration listening only on IP address 127. FEATURE(`accept_unresolvable_domains’)dnl FEATURE(`relay_based_on_MX’)dnl Note You need to be careful with the accept_unresolvable_names feature.Addr=127. then you can uncomment the localhost DAEMON_OPTIONS entry and add another one for the IP address of the NIC on which you wish to accept SMTP traffic. Because sendmail listens on TCP port 25.qxd 2/25/05 10:38 AM Page 344 344 Configuring Linux Mail Servers Chapter 21 become a mail server. However.1:25 0. Name=MTA’) We strongly recommend to comment this one out if you want to protect yourself from spam. . Bigboy the mail server does not accept e-mail relayed from any of the other PCs on your network if they are not in DNS. Determine which NICs sendmail is running on.23Harrison_ch21. See the sixth and next to last lines in the example: dnl dnl dnl dnl dnl dnl dnl … … … dnl dnl dnl dnl dnl dnl dnl dnl This changes sendmail to only listen on the loopback device 127. and not a mail client.0.1.mc to make sendmail listen on all interfaces. DAEMON_OPTIONS(`Port=smtp.

0). this file does not exist in a standard Red Hat/Fedora install.mc file.org’) 4. I usually create two easily identifiable sections in it with all the custom commands I’ve ever added. 5. A General Guide to Using the sendmail.org . Comment this out with a dnl at the beginning.qxd 2/25/05 10:38 AM Page 345 Configuring sendmail 345 3.0. Make sure sendmail is listening on all interfaces (0. The following sections will show you how to define what type of mail it should handle and the various ways this mail can be processed. By default. [root@bigboy tmp]# netstat -an | grep :25 | grep tcp tcp 0 0 0.23Harrison_ch21. you can do this with the restart script from the beginning of the chapter.mc. Regenerate the sendmail.0.0.my-web-site.cf file. and the second section is at the very bottom.org and not destined for this mail server will be forwarded: another-web-site.0.0.mc file can seem jumbled. all mail sent from another-web-site. The mail server doesn’t need a SMART_HOST entry in its sendmail. The first section is near the top where the FEATURE statements usually are. Comment out the SMART_HOST entry in sendmal. dnl define(`SMART_HOST’. Having easily identifiable modifications in the file will make post upgrade reconfiguration much easier. The contents of the relay-domains file should be limited to those domains that can be trusted not to originate spam. and restart sendmail.mc File The sendmail. Again.`mail.0. Sometimes sendmail will archive this file when you do a version upgrade. Here is a sample: dnl ***** Customised section 1 start ***** dnl dnl FEATURE(delay_checks)dnl FEATURE(masquerade_envelope)dnl FEATURE(allmasquerade)dnl FEATURE(masquerade_entire_domain)dnl dnl dnl dnl ***** Customised section 1 end ***** The /etc/mail/relay-domains File The /etc/mail/relay-domains file is used to determine domains from which it will relay mail. To make it less cluttered.0:* LISTEN [root@bigboy tmp]# You have now completed the first phase of converting your Linux server into a sendmail server by enabling it to listen to SMTP traffic on its interfaces.0:25 0. In this case.

the mail server will relay mail only for those PCs on your network that have their e-mail clients configured to use the mail server as their outgoing SMTP mail server. you may find your server being used to relay mail for spam e-mail sites.1.17 192. OK (not ACCEPT). only spam flowing through you. and DISCARD.16 192.1.168.) If you don’t take the precaution of using this feature. my experience has been that control on a per e-mail address basis is much more intuitive via the /etc/mail/virtusertable file.168. such as restricting relaying by IP address or network range and is more commonly used. sendmail will reject. (In Outlook Express. REJECT.0.localdomain localhost 127. two client PCs on your home 192. sendmail assumes it could be either and tries to match both. Despite this.org.168. everyone on your 192.X network. localhost). The second lists the type of action to be taken when mail from these sources or destinations is received.168. Keywords include RELAY. Remember that a server will be considered a part of my-web-site.0.23Harrison_ch21.2.0.X network. The sample file that follows allows relaying for only the server itself (127.0.org only if its IP address can be found in a DNS reverse zone file: localhost. That is to say.qxd 2/25/05 10:38 AM Page 346 346 Configuring Linux Mail Servers Chapter 21 One disadvantage of this file is that it controls mail based on the source domain only.1.168. you set this using: Tools>Accounts>Properties>Servers. The first lists IP addresses and domains from which the mail is coming or going. Configuring the /etc/mail/access file will not stop spam coming to you.org RELAY RELAY RELAY RELAY RELAY RELAY RELAY .1. and everyone passing e-mail through the mail server from servers belonging to my-web-site. The /etc/mail/access File You can make sure that only trusted PCs on your network have the ability to relay mail via your mail server by using the /etc/mail/access file.2 my-web-site. and source domains can be spoofed by spam e-mail servers. All other attempted relayed mail that doesn’t match any of the entries in the /etc/mail/access file. Be sure to run the restart sendmail script from the beginning of the chapter for these changes to take effect. The /etc/mail/access file has two columns. then relay access is fully determined by the /etc/mail/access file.1 192. There is no third column to state whether the IP address or domain is the source or destination of the mail. If you delete /etc/mail/relay-domains. The /etc/mail/access file has more capabilities.

Here is an example.org and another-site then the file would look like this: my-web-site.my-web-site. and the second column lists the local user’s mail box.qxd 2/25/05 10:38 AM Page 347 Configuring sendmail 347 You’ll then have to convert this text file into a sendmail readable database file named /etc/mail/access. If there is no match in the virtusertable file.23Harrison_ch21.org site.org. or a mailing list entry in the /etc/aliases file to which the e-mail should be forwarded. Remember that the relay security features of this file may not work if you don’t have a correctly configured /etc/hosts file. files to determine the ultimate recipient of mail. (Remember each is important. it needs a way of determining whether it is responsible for the mail it receives.db. if this mail server was to accept mail for the domains my-web-site. The first column lists the target e-mail address. It uses the /etc/mail/local-host-names file to do this.org. remember to modify the MX record of the another-webzone file to point to my-web-site.org another-web-site.) . This file has a list of hostnames and domains for which sendmail accepts responsibility. Here are the commands you need: [root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make The sendmail restart script configured at the beginning of the chapter does this for you too. sendmail checks for the full e-mail address in the /etc/aliases file: .org DNS In this case. a remote e-mail address. Primary Mail Exchanger for another-web-site.org. Which User Should Really Receive the Mail? After checking the contents of the virtusertable. For example. The /etc/mail/local-host-names File When sendmail receives mail. The /etc/mail/virtusertable file sendmail checks the aliases The /etc/mail/virtusertable file contains a set of simple instructions on what to do with received mail.org another-web-site. MX 10 mail.

named. In the example that follows.23Harrison_ch21.org webmasters marc sales@another-web-site.org @another-web-site.org finance@my-web-site. lp. If sendmail doesn’t find a duplicate. apache. it assumes the recipient is a regular user on the local server and deposits the mail in their mailbox.qxd 2/25/05 10:38 AM Page 348 348 Configuring Linux Mail Servers Chapter 21 webmaster@another-web-site.org. If there is no match.” After editing the /etc/mail/virtusertable file.org sales@my-web-site. In this case. all other mail to another-web-site.org paul@my-web-site. It then searches for the nickname again in the first column to see if the recipient isn’t on yet another mailing list.org goes to local user marc ☞ sales at my-web-site.org paul paul error:nouser User unknown In this example.org goes to local user (or mailing list) paul ☞ All other users at my-web-site. and when it is all finished. The first column has the mailing list name (sometimes called a virtual mailbox).org @my-web-site. If it finds a match in the first column. you have: They’re in your all-purpose sendmail restart script. shutdown. and so on by system processes will all be sent to user (or mailing list) root. To start. then sendmail assumes the recipient is a regular user on the local server and deposits the mail in their mailbox. you can see that mail sent to users bin. sendmail notes the nickname entry in the second column. mail sent to: webmaster@another-web-site. and the second column has the members of the mailing list separated by commas.db with two commands: [root@bigboy tmp]# cd /etc/mail [root@bigboy mail]# make If these lines look like you’ve seen them before. .org goes to local user (or mailing list) webmasters. If the recipient is a mailing list.com ☞ paul and finance at my-web-site. root is actually an alias for a mailing list consisting of user marc and webmaster@my-web-site. they all get a copy of the e-mail message. you have to convert it into a sendmail-readable database file named /etc/mail/virtusertable.org goes to the sales department at my-othersite. then sendmail goes through the process all over again to determine if any of the members is on yet another list. sendmail searches the first column of the file for a match. daemon.org receive a bounce back message stating “User unknown. The /etc/aliases File You can think of the /etc/aliases file as a mailing list file.

[root@bigboy tmp]# newaliases # Basic system aliases -. paul..org.these MUST be present.org Notice that there are no spaces between the mailing list entries for root: You will get errors if you add spaces. and grandma. you’ll have to convert it into a sendmail readable database file named /etc/aliases. Also. mail sent to root actually goes to user account marc and webmaster@my-web-site. bin: root daemon: root . mailer-daemon: postmaster postmaster: root # General redirections for pseudo accounts...mary ☞ Mail sent to sister: # My family family: family@my-web-site. In this simple mailing list example. . ☞ Mail to directors@my-web-site.brother.org goes to users grandma. # My mailing list file admin-list: “:include:/home/mailings/admin-list” .23Harrison_ch21. abuse: root # trap decode to catch security attacks decode: root # Person who should get root’s mail root: marc. Because aliases can be very useful. after editing this file.org goes to users peter.sister ☞ Mail sent to admin-list gets sent to all the users listed in the file /home/mailings/admin-list.qxd 2/25/05 10:38 AM Page 349 Configuring sendmail 349 Note The default /etc/aliases file installed with Red Hat/Fedora has the last line of this sample commented out with a #.webmaster@my-web-site. here are a few more list examples for your /etc/aliases file.. brother. You may want to delete the comment and change user marc to another user. and mary: # Directors of my SOHO company directors: peter.paul.db. The newaliases command is used to do this.

for example: # Person who should get root’s mail root: webmaster@my-web-site.mc that all outgoing mail originating on Bigboy should appear to be coming from my-web-site. In other words. then you have two choices: user@mysite.com and not ☞ Configure your e-mail client. you made Bigboy the mail server for the domain myweb-site. If either of these are a problem for you. You now have to tell Bigboy in the sendmail configuration file sendmail. Another is that all subscriptions and unsubscriptions have to be done manually by the mailing list administrator. you may find yourself dumping legitimate mail. user root is only needed to update the aliases file. When sendmail sends e-mail to a local user.org. based on the settings in the /etc/hosts file. this forces sendmail to insert the correct fields in the header. to set your e-mail address to user@mysite.my-web-site.mc configuration file and adding some masquerading commands and directives: . To get around this. such as majordomo.com. if not. Despite this.org sendmail Masquerading Explained If you want your mail to appear to come from user@bigboy. One important note about the /etc/aliases file: By default your system uses sendmail to mail system messages to local user root. One is that bounce messages from failed attempts to broadcast go to all users.qxd 2/25/05 10:38 AM Page 350 350 Configuring Linux Mail Servers Chapter 21 The advantage of using mailing list files is that the admin-list file can be a file that trusted users can edit.org. there are some problems with mail reflectors. no matter which server originated the e-mail.) ☞ Set up masquerading to modify the domain name of all traffic originating from and passing trough your mail server. then consider using a mailing list manager. This isn’t terrible. You can solve this by editing your sendmail. (I’ll explain this in the “Configuring Your POP Mail Server” section. such as Outlook Express. try making root have an alias for a user with a fully qualified domain name. If you then use a mail client with a spam mail filtering rule to reject mail with no To: in the header. but you may not want your Web site to be remembered with the word “mail” in front of it.23Harrison_ch21. you may want your mail server to handle all e-mail by assigning a consistent return address to all outgoing mail.com.org. Configuring Masquerading In the DNS configuration. the mail has no To: in the e-mail header.mysite. such as Outlook Express or Evolution. mail will appear to come from mail.

the other recipient sees a cc: to an address he knows instead of one on localhost. ☞ FEATURE masquerade_entire_domain makes sendmail masquerade servers named *my-web-site.com originally. Use this with caution when you are sure you have the necessary authority. ☞ FEATURE masquerade_envelope rewrites the e-mail envelope just as MASQUERADE_AS rewrote the header.org. The allmasquerade statement causes all mail destined for these mailboxes to appear to be destined for users in the domain defined in the MASQUERADE_AS statement.org. If you cc: yourself on an outgoing mail.qxd 2/25/05 10:38 AM Page 351 Configuring sendmail 351 FEATURE(always_add_domain)dnl FEATURE(`masquerade_entire_domain’)dnl FEATURE(`masquerade_envelope’)dnl FEATURE(`allmasquerade’)dnl MASQUERADE_AS(`my-web-site. FEATURE always_add_domain .org and localdomain domains appear to come from the MASQUERADE_AS domain of my-web-site. ☞ always masquerades e-mail addresses.org enters the correct mailbox but sendmail rewrites the To:. If this wasn’t selected. sendmail checks the domain name associated with the IP address of the mail relay client sending the mail to help it determine whether it should do masquerading or not. making the e-mail appear to be sent to peter@myste.org would be masqueraded as my-web-site.org. mail from sales.com would be masqueraded. if MASQUERADE_AS is my-web-site. Note Use FEATURE allmasquerade with caution if your mail server handles e-mail for many different domains and the mailboxes for the users in these domains reside on the mail server.23Harrison_ch21.org and *another-web-site. Using DNS.org by rewriting the e-mail header. In other words. then only servers named myweb-site.’)dnl MASQUERADE_DOMAIN(localhost)dnl MASQUERADE_DOMAIN(localhost. then mail for peter@another-web-site. ☞ The MASQUERADE_DOMAIN directive makes mail relayed via Bigboy from all machines in the another-web-site.org.org and my-othersite.my-web-site.org’)dnl MASQUERADE_DOMAIN(`my-web-site. ☞ FEATURE allmasquerade makes sendmail rewrite both recipient addresses and sender addresses relative to the local machine.localdomain.org as my-web-site.localdomain)dnl The result is that: ☞ The MASQUERADE_AS directive makes all mail originating on Bigboy appear to come from a server within the domain my-web-site.org and you use allmasquerade. In other words. even if the mail is sent from a user on the mail server to another user on the same mail server.

All e-mail correspondence then has a uniform e-mail address format that complies with the company’s brand marketing policies. it is detrimental to e-mail delivery. The e-mail envelope contains the To: and From: used by mail servers for protocol negotiation. not only because apache’s . send mail as user apache too. but the program inserts its own address as sender’s address.qxd 2/25/05 10:38 AM Page 352 352 Configuring Linux Mail Servers Chapter 21 Masquerading is an important part of any mail server configuration as it enables systems administrators to use multiple outbound mail servers. When you choose Reply or Reply All in Outlook Express. Testing Masquerading The best way of testing masquerading from the Linux command line is to use the mail -v username command. It is the envelope’s From: that is used when e-mail rejection messages are sent between mail servers. perhaps you bought a program for your SOHO office that sends out notifications to your staff. however. consider the To: and From: statements as the e-mail header. to fake the envelope. user root will not be masqueraded. For example. Web-based CGI scripts tend to run as user apache and. each providing only the global domain name for a company and not the fully qualified domain name of the server itself. Using sendmail to Change the Sender’s E-mail Address Sometimes masquerading isn’t enough.mc. as spammers often do.23Harrison_ch21. You can comment this out if you like with a “dnl” at the beginning of the line and running the sendmail start script. the program automatically uses the To: and From: in the header. You should also tail the /var/log/maillog file to verify that the masquerading is operating correctly and check the envelope and header of test e-mail received by test e-mail accounts. not that of the IT person. therefore. Other Masquerading Notes By default. Note E-mail clients. I have noticed that sendmail -v username ignores masquerading altogether. such as Outlook Express. To remove this restriction use the EXPOSED_USER(`root’)dnl command in /etc/mail/sendmail. It is easy to fake the header. Often you won’t want this. At times you may need to change not only the domain of the sender but also the username portion of the sender’s email address.

org 4.mc file to activate the FEATURE(`genericstable’. Here are a few tests and techniques you can use to fix some of the most common problems.org bigboy. you can change both the domain and username on a caseby-case basis using the genericstable feature: 1.qxd 2/25/05 10:38 AM Page 353 Configuring sendmail 353 e-mail address may not be suitable. Your e-mails from linux-username should now appear to come from username@new-domain. The format of the file is: Here is an example: alert peter apache security-alert@my-web-site. Troubleshooting sendmail When sendmail doesn’t appear to work correctly.org mailer@my-web-site. It then looks at the /etc/mail/genericstable file for an individual e-mail linux-username username@new-domain. If your virtusertable file allows e-mail to only predefined users.org 3. actually exists as a real user. Make sure the file includes your server’s canonical domain name.db’)dnl GENERICS_DOMAIN_FILE(`/etc/mail/generics-domains’)dnl 2. . Create a /etc/mail/generics-domains file that is just a list of all the domains that should be inspected. First sendmail searches the /etc/mail/generics-domains file for a list of domains to reverse map.23Harrison_ch21.`hash -o /etc/mail/genericstable.my-web-site.org urgent-message@my-web-site.com.org another-web-site. Run the sendmail restart script from the beginning of the chapter and then test.com address from a matching domain. Add these statements to your feature: /etc/mail/sendmail. and your valid e-mail may be classified as being spam. Create your /etc/mail/genericstable file.1 </dev/null Here is a sample /etc/mail/generics-domains file: my-web-site. With sendmail. which you can obtain using the command: sendmail -bt -d0. you can test it in a number of ways. or source e-mail address. but also because some anti-spam programs check to ensure that the From:. then queries about the apache user will fail.

org 25 Trying mail. Recipient ok 5. Test TCP Connectivity You can also mimic a full mail session using TELNET to make sure everything is working correctly. Escape character is ‘^]’.my-web-site. Failure could also be caused by the sendmail process being stopped.. the cause is probably a typographical error.comcast. Inform the mail server from which the test message is coming with the MAIL FROM: statement: MAIL FROM:sender@another-web-site.org Hello c-24-4-97110.my-web-site.org. Using the RCPT sage is going: TO: statement. It is best to test this from both inside your network and from the Internet.0 sender@another-web-site.my-web-site. If you get a “500 Command not recognized” error message along the way.1.org.” covers troubleshooting with TELNET.org 250 2. You should get a response with a 220 status code. Connected to mail. or port forwarding rules to your mail server. Sender ok 4.4. [root@bigboy tmp]# telnet mail.org.. Chapter 4. Lack of connectivity could be caused by a firewall with incorrect permit.net [24. NAT..my-web-site. Type in your text message.. 3.org ESMTP server ready 2. ending it with a single period on the last line. “Simple Network Troubleshooting. TELNET to the mail server on port 25. 220 mail.1..client. tell the mail server to whom the test mes- RCPT TO: user@my-web-site.my-web-site. You should receive a message with a successful status 250 code at the beginning of the response. For example: . Type the string “subject:” then type a subject.org 250 mail. pleased to meet you. Use the hello command to tell the mail server the domain to which you belong. Prepare the mail server to receive data with the DATA statement: DATA 354 Enter mail. hello another-web-site.org 250 2. Follow these steps carefully: 1.110].” on a line by itself 6.23Harrison_ch21. end with “.qxd 2/25/05 10:38 AM Page 354 354 Configuring Linux Mail Servers Chapter 21 Test TCP Connectivity The very first step is to determine whether your mail server is accessible on the sendmail SMTP TCP port 25..5 user@my-web-site.org.97.

12.0.23Harrison_ch21. Use the QUIT command to end the session.my-web-site.0 iA75r9si017840 Message accepted for delivery 7. Open two TELNET. [root@bigboy tmp]# Now verify that the intended recipient received the message. QUIT 221 2. .cf /etc/mail/sendmail.org closing connection Connection closed by foreign host.5 supports version 10.mc file [root@bigboy mail]# m4 /etc/mail/sendmail.cf file is version 0 No local mailer defined QueueDirectory (Q) option must be set [root@bigboy mail]# ☞ Sample errors when processing the sendmail.qxd 2/25/05 10:38 AM Page 355 Configuring sendmail 355 Subject: Test Message Testing sendmail interactively .0. and check the system logs for any mail application errors.cf file is out of date: sendmail 8. SSH.m4: No such file or directory [root@bigboy mail]# ☞ Sample errors when restarting sendmail [root@bigboy mail]# service sendmail restart Shutting down sendmail: [ OK ] Shutting down sm-client: [FAILED] Starting sendmail: 554 5. or console windows. If they are not. The /var/log/maillog File Because sendmail writes all its status messages in the /var/log/maillog file. These must be installed. 250 2. Work in one of them and monitor the sendmail status output in the other using the command: [root@bigboy tmp]# tail -f /var/log/maillog Common Errors Due to Incomplete RPM Installation Both the newaliases and m4 commands require the sendmail-cf and m4 RPM packages.mc:8: m4: Cannot open /usr/share/sendmailcf/m4/cf.0 mail. always monitor this file whenever you are doing changes.0. you’ll get errors when running various sendmail related commands: ☞ Sample errors when running newaliases [root@bigboy mail]# newaliases Warning: .0 No local mailer defined .mc > /etc/mail/sendmail.

localdomain localhost Unfortunately in this configuration.02(BFLITEMAIL4A).com or . Enter some text after issuing the command and end your message with a single period all by itself on the last line. As mentioned.com domain and would ignore the security features of the access and relaydomains files I’ll describe later.0.another-web-site. via esmtp. It will then default to sending emails in which the domain is localhost. a poorly configured /etc/hosts file can make mail sent from your server to the outside world appear as if it came from users at localhost.0.org. Connecting to mail. example@another-web-site. 220 ltmail.qxd 2/25/05 10:38 AM Page 356 356 Configuring Linux Mail Servers Chapter 21 554 5.org.1 bigboy localhost. sendmail. Fedora inserts the host name of the server between the 127.1 (Wrong!!!) my-web-site. The /etc/hosts file is also important for configuring mail relay.localdomain 250-mx.120.my-web-site.org Hello [67..org LiteMail v3.localdomain and not bigboy.com: 127.0 QueueDirectory (Q) option must be set [FAILED] Starting sm-client: [ OK ] [root@bigboy mail]# If these errors occur.localdomain.localdomain localhost The server would therefore be open to relay all mail from any .. and sendmail-cf RPM By default.0.another-web-site..23Harrison_ch21.106].0. You can create problems if you fail to place the server name in the FDQN for 127.org test text test text .221. Use the sendmail program to send a sample e-mail to someone in verbose mode.net. pleased to meet you 250 HELP . such as . Here sendmail thinks that the server’s FDQN was my-site and that the domain was all of .0. Incorrectly Configured /etc/hosts Files m4.another-website..1 and the localhost entries in /etc/hosts: 127. 05 Oct 2002 06:48:44 -0400 >>> EHLO localhost.org. Sat.0. make sure your packages are installed correctly. which it will identify as being invalid because there is no extension at the end. for example: [root@bigboy tmp]# sendmail -v example@another-web-site. sendmail will think that the server’s FQDN is Bigboy.org localhost.0.0.1 entry.0.

org> 250 <example@another-web-site. it is therefore an illegal Internet domain. Consider an example: Mail sent from computer PC1 to PC2 appears to come from a user at localhost..) [root@bigboy tmp]# sendmail –v root WARNING: local host name (bigboy) is not qualified. 250 Message accepted for delivery example@another-web-site. sleeping for retry . >>> QUIT [root@bigboy tmp]# Localhost.” on a line by itself >>> .org. The rejected e-mail is returned to localhost. too. An error in the script could cause this type of message. Another set of telltale errors caused by the same problem can be generated when trying to send mail to a user (the example uses root) or creating a new alias database file.org.localdomain is the domain that all computers use to refer to themselves. Recipient Ok >>> DATA 354 Enter mail.localdomain and thinks that the rejected e-mail should be sent to a user on PC2 that may not exist. fix $j in config file [root@bigboy tmp]# An accompanying error in /var/log/maillog log file looks like this: Oct 16 10:23:58 bigboy sendmail[2582]: My unqualified host name (bigboy) unknown. PC2 sees that the mail originated from localhost..localdomain on PC1 and is rejected.localdomain> 250 <root@localhost.. You end up with an error in /var/log/maillog: Oct 16 10:20:04 bigboy sendmail[2500]: g9GHK3iQ002500: SYSERR(root): savemail: cannot save rejected email anywhere Oct 16 10:20:04 bigboy sendmail[2500]: g9GHK3iQ002500: Losing .23Harrison_ch21. Sent (Message accepted for delivery) Closing connection to mail.localdomain>. such as a script based on the PERL module Mail::Audit. end with “..org>.. (I’ll explain the newaliases command later.localdomain..another-web-site. Sender Ok >>> RCPT To:<example@another-web-site./qfg9GHK3iQ002500: savemail panic You may also get this error if you are using a spam prevention program.qxd 2/25/05 10:38 AM Page 357 Configuring sendmail 357 >>> MAIL From:<root@localhost. fix $j in config file [root@bigboy tmp]# newaliases WARNING: local host name (bigboy) is not qualified.

html)”’. and in some cases dangerous when they contain viruses and worms.easynet.see http://www. `”550 Email rejected due to sending server misconfiguration . Here are some sample entries you can add to your /etc/sendmail.`sbl.blackholes. `’)dnl ☞ The Open Relay Database: An open mail relay list.org’.see http://www. Some spam blacklists simply try to determine whether the e-mail is coming from a legitimate IP address.nl’.spamhaus. but in all cases they rely heavily on complaints from spam victims. others are managed by public companies. In some cases.ordb. `proxies.org/’)dnl Be sure to visit the URLs listed to learn more about the individual services.1 ACCESS DENIED to OPEN PROXY SERVER “$&{client_name}” by easynet.mc file. you can use your mail server to combat spam. Some are maintained by volunteers.see http://spamhaus. FEATURE(`dnsbl’.nl DNSBL (http://proxies. Fortunately. `relays.net/bl. if the spammer uses a public data center.`Rejected . `”550 5. FEATURE(`dnsbl’.ordb.shtml”’) ☞ Spamhaus: A spammer blacklist. Using Public Spam Blacklists with sendmail There are many publicly available lists of known open mail relay servers and spam generating mail servers on the Internet. .rfc-ignorant.7. it is possible that their activities could also cause the IP addresses of legitimate e-mailers to be black listed too. `”450 Mail from “ $`’&{client_addr} “ refused . time consuming to delete.qxd 2/25/05 10:38 AM Page 358 358 Configuring Linux Mail Servers Chapter 21 FIGHTING SPAM Unsolicited commercial e-mail (UCE or spam) can be annoying. It is hoped that these legitimate users will pressure the data center’s management to evict the spamming customer.org/”’) ☞ Easynet: An open proxy list.rfc-ignorant. FEATURE(`dnsbl’.easynet.net’. the blacklists include not only the offending IP address but also the entire subnet or network block to which it belongs. The IP addresses of offenders usually remain on the list for six months to two years. to provide additional pressure on the spammers. FEATURE(`dnsbl’. FEATURE(`dnsbl’.org’.blackholes. You can configure sendmail to use its dnsbl feature to both query these lists and reject the mail if a match is found.org/faq/\#why_rejected”’)dnl ☞ Spamcop: A spammer blacklist. Also. `bl. `ipwhois. Rejected for bad WHOIS info on IP of your SMTP server .nl/errors.org’.`”550 Mail from “ $&{client_addr} “ refused.spamcop.see http://spamcop.23Harrison_ch21. they should all be on one line: ☞ RFC-Ignorant: A valid IP address checker. This prevents the spammers from easily switching their servers’ IP addresses to the next available ones on their networks.

and restart spamassassin after booting: [root@bigboy tmp]# service spamassassin start [root@bigboy tmp]# service spamassassin stop [root@bigboy tmp]# service spamassassin restart Configuring procmail for spamassassin The /etc/procmailrc file is used by procmail to determine the procmail helper programs that should be used to filter mail. Downloading and Installing spamassassin Most Red Hat and Fedora Linux software products are available in the RPM format.qxd 2/25/05 10:38 AM Page 359 Fighting Spam 359 spamassassin Once sendmail receives an e-mail message.60-2. remember that the filename usually starts with the software package name and is followed by a version number.rpm. see Chapter 6. as in spamassassin-2. spamassassin doesn’t delete spam. This file isn’t created by default. The most commonly used filter is spamassassin.i386. it hands the message over to procmail. When searching for the RPMs. You can then configure the e-mail filter rules in Outlook Express or any other mail client to either delete the suspect message or store it in a special Spam folder. .rc. You can make procmail temporarily hand over control to another program.rc /etc/procmailrc This file forces all mail arriving for your mail server’s users through Spamassasin. stop. such as a spam filter. (For help downloading. Copy the template to the /etc directory: [root@bigboy tmp]# cp /etc/mail/spamassassin/spamassassin-spamc. spamassassin has a template you can use called /etc/mail/spamassassin/ spamassassin-spamc. which is the application that actually places the e-mail in user mailboxes on the mail server.) Starting spamassassin You can use the boot: chkconfig command to get spamassassin configured to start at [root@bigboy tmp]# chkconfig --level 35 spamassassin on To start. it merely adds the word “spam” to the beginning of the subject line of suspected spam e-mails.23Harrison_ch21.

0=No) skip_rbl_checks 0 use_razor2 1 use_dcc 1 use_pyzor 1 # Mail using languages used in these country codes will not be marked # as being possibly spam in a foreign language. # .23Harrison_ch21. 0=No) use_bayes 1 # Enable Bayes auto-learning (1=Yes.cf (see spamassassin(1) for details) # # How many hits before a message is considered spam.0 # Whether to change the subject of suspected spam (1=Yes. The lower the # number the more sensitive it is. required_hits 5. You can customize this fully meet your needs: file is named /etc/mail/spamassassin/ commented sample configuration file to ################################################################### # See ‘perldoc Mail::SpamAssassin::Conf’ for # details of what can be adjusted.cf.spamassassin/user_prefs. 0=No) report_safe 1 # Use terse version of the spam report (1=Yes.english ok_languages en . 0=No) auto_learn 1 # Enable or disable network checks (1=Yes. 0=No) use_terse_report 0 # Enable the Bayes system (1=Yes. ################################################################### # # These values can be overridden by editing # ~/.qxd 2/25/05 10:38 AM Page 360 360 Configuring Linux Mail Servers Chapter 21 Configuring spamassassin The spamassassin configuration local. 0=No) rewrite_subject 1 # Text to prepend to subject if rewrite_subject is used subject_tag *****SPAM***** # Encapsulate spam in an attachment (1=Yes.

Place an executable version of the script in your home directory and modify the script’s $FILEPATH variable point to your home directory. because sendmail always checks the . By default. This is fairly simple to do.cpan. The sendmail program then looks for the filename in the directory /etc/smrsh and executes it.forward file and place an entry in /etc/smrsh. Startup spamassassin The final steps are to configure start it. and file mail-filter. which specifies the subjects and e-mail addresses to accept. 4. Update file mail-filter. which specifies those to reject.forward file in your home directory for the name of this script. PERL doesn’t come with modules that are able to check e-mail headers and envelopes so you have to download them from CPAN (www. Install PERL and the PERL modules you downloaded from CPAN.23Harrison_ch21. spamassassin to start on booting and then to [root@bigboy tmp]# chkconfig spamassassin on [root@bigboy tmp]# service spamassassin start Starting spamd: [ OK ] [root@bigboy tmp]# A Simple PERL Script to Help Stop Spam Blacklists won’t stop everything. ok_locales en Be sure to restart spamassassin for your changes to take effect. The most important modules are: ☞ ☞ ☞ ☞ MailTools IO-Stringy MIME-tools Mail-Audit I have written a script called mail-filter. .reject. 2.pl that effectively filters out spam e-mail for my home system.qxd 2/25/05 10:38 AM Page 361 Fighting Spam 361 # Mail using locales used in these country codes will not be marked # as being possibly spam in a foreign language.org). A few steps are required to make the script work: 1.accept. but you can limit the amount of unsolicited spam you receive by writing a small script to intercept your mail before it is written to your mailbox. Update your . 3.

and Configurations. If you want to retrieve the mail from your Linux box’s user account using a mail client such as Evolution. It then denies everything else. then you have a few more steps. see Appendix II. the chkconfig command also immediately activates application. and you will probably have to download it from rpmfind.i386. Microsoft Outlook or Outlook Express. once again use chkconfig: [root@bigboy tmp]# chkconfig pop3 off Remember to restart the POP mail process every time you make a change to the configuration files to ensure the changes take effect on the running process.23Harrison_ch21. but I have found the IMAP-2002 RPMs found on rpmfind. To configure POP mail to start at boot. but sendmail just handles mail sent to your my-web-site.net. Remember that the filename is probably similar to imap-2002d-3. “Codes. It isn’t yet a part of the Fedora RPM set. For a simple script with instructions on how to install the PERL modules. [root@bigboy tmp]# chkconfig pop3 on To stop POP mail after booting. As with all xinetd-controlled programs. therefore.net and featured in this section much more intuitive to use for the SOHO environment. Installing Your POP Mail Server You need to install the imap RPM that contains the POP server software. Starting Your POP Mail Server POP mail is started by xinetd. You’ll also have to make your Linux box a POP mail server. . Scripts.” CONFIGURING YOUR POP MAIL SERVER Each user on your Linux box will get mail sent to their account’s mail folder. Fedora Linux comes with its Cyrus IMAP/POP server RPM package.qxd 2/25/05 10:38 AM Page 362 362 Configuring Linux Mail Servers Chapter 21 Mail-filter first rejects all e-mail based on the reject file and then accepts all mail found in the accept file.rpm.org domain. you have to use the chkconfig command to make sure xinetd starts up on booting.

both users will get sent to the Linux user account john by default. so firewall rules may need to be altered as well. then you may need to modify your virtusertable file. If the users insist on overlapping names. You have two options for a solution: ☞ Make the user part of the e-mail address different—john1@my-website. use /etc/xinetd. set your SMTP mail server to be the IP address/domain name of your Linux mail server. for example—and create Linux accounts john1 and john2. or port . you may want to configure POP to work over an encrypted data channel. NAT. For example. Use your Linux user username and password when prompted.org. Encrypted POP runs on TCP port 995. set your POP mail server to be the IP address of your Linux mail server. Lack of connectivity could be caused by a firewall with incorrect permit.org). Most POP clients support secure POP. The POP configuration in Outlook Express for each user should retrieve their mail via POP using john1 and john2. Troubleshooting POP Mail The very first troubleshooting step is to determine whether your POP server is accessible on the POP TCP port 110 or the secure POP port of 995.d/ipop3. Configuring Secure POP Mail If you need to access your e-mail from the mail server via the Internet or some other insecure location.org entries to account john2. To configure POP Mail.d/pop3s file instead of /etc/xinetd.23Harrison_ch21. With this trick you’ll be able to handle many users belonging to multiple domains without many address overlap problems. respectively.org) and John Brown (john@another-web-site. such as John Smith (john@my-web-site. Next.org and john2@another-web-site. ☞ Create the user accounts john1 and john2 and point virtusertable entries for john@my-web-site.org to account john1 and point john@another-web-site.qxd 2/25/05 10:38 AM Page 363 Configuring Your POP Mail Server 363 How to Configure Your Windows Mail Programs All your POP e-mail accounts are really only regular Linux user accounts in which sendmail has deposited mail. Windows configures it in the Advanced menu of the Outlook Express Account Configuration window. For this. You can now configure your e-mail client such as Outlook Express to use your new POP/SMTP mail server quite easily. How to Handle Overlapping E-mail Addresses If you have user overlap.

16. (Troubleshooting TCP with TELNET is covered in Chapter 4.org [172.16.1.” discusses many Linux-based tools that you can use to track the health of your Linux server.) Linux status messages are logged to the file /var/log/messages.16.23Harrison_ch21.103 Aug 11 23:20:33 bigboy ipop3d[18693]: pop3s SSL service init from 172. Check your configuration files if it fails to do so.org [172.103] nmsgs=0/0 Aug 11 23:20:52 bigboy ipop3d[18694]: Logout user=labmanager host=17216-1-103.103] nmsgs=0 ndele=0 Aug 11 23:20:40 bigboy xinetd[18690]: EXIT: pop3s pid=18693 duration=7(sec) Aug 11 23:20:52 bigboy xinetd[18690]: START: pop3s pid=18694 from=172.1. Chapter 22. “Monitoring Server Performance. (Linux logging is covered in Chapter 5.1. Use it to make sure all your files are loaded when you start xinetd. “Troubleshooting with syslog.1.103 Aug 11 23:20:52 bigboy ipop3d[18694]: Login user=labmanager host=17216-1-103.1.”) Aug 11 23:20:33 bigboy xinetd[18690]: START: pop3s pid=18693 from=172. your Web site won’t seem complete. It needs to be maintained to reduce the risk of failure and monitored to help detect potential problems. This example starts xinetd and makes a successful secure POP query from a remote POP client.1.org [172.my-web-site.qxd 2/25/05 10:38 AM Page 364 364 Configuring Linux Mail Servers Chapter 21 forwarding rules to your server.1. A fully functioning Web site is just the beginning. and you need to plan its configuration carefully to make it a seamless part of the Web experience of your visitors.103 Aug 11 23:20:40 bigboy ipop3d[18693]: Login user=labmanager host=17216-1-103. .1.16. Without it.103] nmsgs=0 ndele=0 Aug 11 23:20:52 bigboy xinetd[18690]: EXIT: pop3s pid=18694 duration=0(sec) CONCLUSION E-mail is an important part of any Web site.16.my-web-site.16.103] nmsgs=0/0 Aug 11 23:20:40 bigboy ipop3d[18693]: Logout user=labmanager host=17216-1-103. Test this from both inside your network and from the Internet.16. Failure could also be caused by the xinetd process being stopped or the configuration files being disabled.org [172.my-web-site.103 Aug 11 23:20:52 bigboy ipop3d[18694]: pop3s SSL service init from 172.16.my-web-site.

For ease of use. The tree starts from a root node. which then descends through branches and leaves that each add their own reference 365 . SNMP Most servers. including MRTG.qxd 2/25/05 10:14 AM Page 365 C H A P T E R 22 Monitoring Server Performance IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ SNMP MRTG Webalizer The top Command The vmstat Command The free Utility Conclusion You can monitor your system’s Web performance quite easily with graphical Linux tools. equipment vendors provide Management Information Base (MIB) files for their devices that define the functions of the OIDs they contain. That’s a lot of new terms to digest in two sentences. which is based on SNMP and monitors raw network traffic. which tracks Web site hits. so take a moment to look more closely.24Harrison_ch22. routers and firewalls keep their operational statistics in object identifiers (OIDs) that you can remotely retrieve via the Simple Network Management Protocol (SNMP). You’ll learn how to use several in this chapter. and Webalizer. OIDs and MIBs OIDs are arranged in a structure of management information (SMI) tree defined by the SNMP standard.

6.1.1.1 shows an OID structure in which the path to the enterprises OID branch passes through the org. root node ccitt (0) iso (1) joint (2) org (3) dod (6) internet (1) directory (1) mgmt (2) experimental (3) private (4) enterprises (1) Figure 22. therefore. dod.24Harrison_ch22. . Figure 22.1 The SNMP OID structure. and private branches first.3. 1.4.qxd 2/25/05 10:14 AM Page 366 366 Monitoring Server Performance Chapter 22 value to the path separated by a period. The OID path for enterprises is. internet.

Remember.1.1.1. and description.3 1.6.4. just the MIB file in which it was found and the alias within the MIB. For example: SNMPv2-MIB::sysUpTime.4.6.9.1. which is named dod. For example.1.3. This MIB information is very useful for SNMP management programs. and they must also provide information in universally accepted OIDs for ease of manageability.1 1.6.6. only the OID value at the very tip of a branch.6.109. For example.4.1.3.1.1. for now. Think of OIDs like the directory structure on a hard disk. The Linux snmpget command outputs the value of a single leaf. and the very last value at the tip (the leaf) correlates to a file containing data.9.6.1. the SMI org MIB defines all the topmost OIDs found at the next layer.3.6.2 1.3.1.9.3.3.5. I’ll discuss these commands later.1 by substituting the branch name (enterprises) for its OID numbers (1.1 OIDs and Their Equivalent MIBs MIB org dod internet directory mgmt experimental private enterprises You can refer to an OID by substituting the values in a branch with one of these more readable MIB aliases. Table 22. and private branches.3.1).1 OID 1.1.1. NIC .0 Here the OID value was found in the SNMPv2-MIB file and occupies position zero in the sysUpTime alias. mgmt.4 1.1.3.5 as enterprises.3 1.1 1. the internet MIB under dod defines the function of the topmost OIDs in the directory.9.qxd 2/25/05 10:14 AM Page 367 SNMP 367 Management Information Bases (MIBs) are text definitions of each of the OID branches.24Harrison_ch22. Each branch is equivalent to a subdirectory. Equipment manufacturers are usually assigned their own dedicated OID branch under the enterprises MIB.6 1.109. type.3. For example.1. all you need to know is that the command output frequently doesn’t list the entire OID.1. Table 22. experimental.1 shows how some commonly used OIDs map to their MIB definitions.1. you can reference the OID 1. enabling you to click on an OID and see its value. and the snmpwalk command provides the values of all leaves under a branch.6. the —leaf— actually has a readable value.

but a memory use value on a customized processor card may be defined in a MIB under the manufacturer’s own OID branch.24Harrison_ch22. you should change it from this easy-to-guess value whenever possible. version 1 was designed to be a protocol to provide device statistics and error reporting without consuming a lot of system resources. or community string. as usernames and passwords have been introduced. ☞ SNMP Version 3: This version provides greater security and remote configuration capabilities than its predecessors. expanded the number of supported error codes. and had the ability to do bulk queries that more efficiently loaded response packets with data. Security was limited to community strings and access controls based on the IP address of the querying server. SNMP Community Strings As a security measure. SNMP v2c is backward compatible with version 1. you need to know the SNMP password. The Read/Write or “set” community string is for not only viewing statistics and system parameters but also for updating the parameters. In many cases the Read Only community string or password is set to the word “public”. Access isn’t limited to a single community string for read-only and read/write access. Data communication wasn’t encrypted. Views of OIDs in a MIB can also be limited on a per-user basis. often referred to as v2c. Remember their differences. because you will need to specify the version number when doing SNMP queries. Support for encrypted SNMP data transfer and transfer error detection is also provided. . the most commonly used ones are the Read Only or “get” community string that only provides access for viewing statistics and system parameters. increased the size of counters used to track data. to query OIDs.qxd 2/25/05 10:14 AM Page 368 368 Monitoring Server Performance Chapter 22 interface data throughput values must always be placed in a predefined location in the general tree. There are a number of types of community strings. ☞ SNMP Version 2: The second version of SNMP. SNMP Versions There are currently three versions of SNMP: ☞ SNMP Version 1: The first version of SNMP to be implemented.

see Chapter 6. remember that the SNMP utility tools RPM’s filename usually starts with net-snmp-utils. [root@bigboy tmp]# Upon inspecting the output of the snmpwalk command. as in net-snmp-utils5. They also require you to specify the version of the SNMP query to be used with a -v 1. “Installing RPM Software... -v 2c.3 = STRING: 0:b0:d0:46:32:71 . IF-MIB::ifDescr.”) SNMP Utilities Command Syntax The SNMP utility tools package installs a number of new commands on your system for doing SNMP queries. (If you need an installation refresher. other servers or even itself using MRTG or any other tool. [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost interface .. but it does require a number of detailed steps. respectively.. most notably snmpget for individual OIDs and snmpwalk for obtaining the contents of an entire MIB. The first argument is the name or IP address of the target device and all other arguments list the MIBs to be queried.1 = STRING: IF-MIB::ifPhysAddress..rpm. Installing SNMP Utilities on a Linux Server If you intend to use your Linux box to query your network devices.1 = STRING: lo IF-MIB::ifDescr. you can see that the second interface seems to have the name eth0 and the MAC address 0:9:5b:2f:9e:d5. .1.2 = STRING: 0:9:5b:2f:9e:d5 IF-MIB::ifPhysAddress. and 3.. which is followed by a version number.. Both commands require you to specify the community string with a -c operator. This example gets all the values in the interface MIB of the local server using SNMP version 1 and the community string of craz33guy.24Harrison_ch22. 2. you need to have the SNMP utility tools package net-snmp-utils installed.i386. You can now retrieve the individual MAC address using the snmpget command: [root@bigboy tmp]# snmpget -v 1 -c const1payted localhost ifPhysAddress.qxd 2/25/05 10:14 AM Page 369 SNMP 369 Doing SNMP Queries Configuring SNMP on a server isn’t hard.. ..2 = STRING: eth0 IF-MIB::ifDescr.2 = STRING: 0:9:5b:2f:9e:d5 [root@bigboy tmp]# . IF-MIB::ifPhysAddress.... or -v 3 operator for versions 1.1-2. . When searching for the file.3 = STRING: eth1 .2 IF-MIB::ifPhysAddress.

The version of the configuration file that comes with net-snmp is quite complicated. Enter this line in the new configuration file to set the Read Only community string to craz33guy: rocommunity craz33guy 3.4. This package uses a configuration file named /etc/snmp/snmpd. Configure Linux to start SNMP services on each reboot with the chkconfig command: [root@bigboy root]# chkconfig snmpd on [root@bigboy root]# 4.119. Configuring SNMP on a Linux Server By default Fedora. installs the net-snmp package as its SNMP server product.old [root@bigboy snmp]# vi snmpd. Here is an example: 1.255 Mask:255.255. I suggest archiving it and using a much simpler version with only a single line containing the keyword rocommunity followed by the community string.240 .. [root@bigboy tmp]# You’ll now see how you can configure SNMP on your Linux server to achieve these results. Start SNMP to load the current configuration file: [root@bigboy root]# service snmpd start Starting snmpd: [ OK ] [root@bigboy root]# 5.10.18-14 #1 Wed Sep 4 11:57:57 EDT 2002 i586 .244 Bcast:216. Save the old configuration file: [root@bigboy tmp]# cd /etc/snmp/ [root@bigboy snmp]# mv snmpd.0 = STRING: Linux bigboy 2.conf snmpd.conf... Test whether SNMP can read the snmpwalk command: system and interface MIBs using the [root@bigboy snmp]# snmpwalk -v 1 -c craz33guy localhost system SNMPv2-MIB::sysDescr.qxd 2/25/05 10:14 AM Page 370 370 Monitoring Server Performance Chapter 22 You can confirm this information using the ifconfig command for interface eth0..255.24Harrison_ch22.10. the very first line shows a matching MAC address: [root@bigboy tmp]# ifconfig -a eth0 eth0 Link encap:Ethernet HWaddr 00:09:5B:2F:9E:D5 inet addr:216.conf 2.119. .conf in which the community strings and other parameters may be set.

. [root@bigboy snmp]# snmpwalk -v 1 -c craz33guy localhost interface IF-MIB::ifNumber.. and -v 3 switches for specifying the SNMP version to be used for queries.. You can also configure your server to use the TCP wrappers feature outlined in Appendix I.1 = INTEGER: 1 IF-MIB::ifIndex. -v 2c. to create online graphs of your traffic flows. such as the Internet. versions 1.168.10 SNMPv2-MIB::sysUpTime.qxd 2/25/05 10:14 AM Page 371 SNMP 371 SNMPv2-MIB::sysObjectID...2 = INTEGER: 2 IF-MIB::ifIndex. All you need is the IP address and SNMP Read Only string and you’ll be able to get similar results.0 = STRING: root@localhost SNMPv2-MIB::sysName. .3 = STRING: eth0 .0 = INTEGER: 3 IF-MIB::ifIndex. you can configure SNMP statistics gathering software. [root@bigboy snmp]# Now that you know SNMP is working correctly on your Linux server.0 = STRING: bigboy ...1. [root@bigboy snmp]# snmpwalk -v 1 -c chir1qui 192. .0 = Timeticks: (425) 0:00:04...conf file.25 SNMPv2-MIB::sysContact.24Harrison_ch22. 2 and 3. such as MRTG.1 = STRING: lo IF-MIB::ifDescr. You should also make sure that you use all reasonable security measures to allow queries only from trusted IP addresses either via a firewall or the SNMP security features available in the snmp. .0 = OID: NET-SNMP-MIB::netSnmpAgentOIDs.3 = INTEGER: 3 IF-MIB::ifDescr.1.168. Basic SNMP Security The most commonly supported versions of SNMP don’t encrypt your community string password so you shouldn’t do queries over insecure networks. you were polling localhost. SNMP on Other Devices In the example.2 = STRING: wlan0 IF-MIB::ifDescr.. The Linux snmpwalk and snmpget commands have -v 1.1.1 interface Different SNMP Versions There are currently three versions of SNMP... Here is an example of a query of a device with an IP address of 192. Always make sure you are using the correct one. You can poll any SNMP-aware network device that has SNMP enabled.. .

conf file can support limiting MIB access to trusted hosts and networks.1.name ======== local network_1 network_2 ## MIB.model ========= v1 v1 v1 sec.” to limit access to specific servers without the need of a firewall. 172.16. The example that follows is a modification of the less restrictive section. the snmpd. In case you need it.1.168.level prefix read ========== ======= ========= ========= ====== ==== write ===== MyROGroup “” v1 noauth 0 all-mibs none In the example: ☞ Only three networks (localhost.1 mask ==== 80 ## ## notif ## ===== access none MIB group.name ## ============== view all-mibs incl/excl ========= included MIB. ☞ Each network is matched to a group called MyROGroup using SNMP version 1. The snmpd.0/24 192.0/24 community ========= craz33guy craz33guy craz33guy ## ## group group group Access.name ================= MyROGroup MyROGroup MyROGroup sec.conf file has two security sections.16.model sec. ## ## com2sec com2sec com2sec sec.0/24. and 192.group.1.subtree =========== .24Harrison_ch22.0/24) are allowed to access the server with the craz33guy community string.name ======== local network_1 network_2 source ====== localhost 172. You will have to comment out the more restrictive statements at the top of the file for it to work correctly.view.1.qxd 2/25/05 10:14 AM Page 372 372 Monitoring Server Performance Chapter 22 “Miscellaneous Linux Topics. and all the MIBs on the server are defined by the view named all-mibs. a section with very restrictive access sits at the top of the file and is immediately followed by a less restrictive section. .name context sec.16. ☞ An access statement ensures that only the defined networks have read only access to all the MIBs.

Figure 22. The product is available from the MRTG Web site (www. ☞ Any TCP wrappers configuration on your SNMP target machine allows SNMP queries from your SNMP client. the snmpd. These precautions are probably unnecessary in a home environment where access is generally limited to devices on the home network by an NAT firewall. You can easily create graphs of traffic flow statistics through your home network’s firewall/router or even your Linux box’s NIC cards using MRTG. Generally speaking in a home environment protected by NAT your TCP wrappers files (/etc/hosts.2 shows a sample MRTG graph.qxd 2/25/05 10:14 AM Page 373 MRTG 373 ☞ Modification of the MIBs via SNMP is denied because the word “none” is in the write section of the access statement.24Harrison_ch22. ☞ Your SNMP security policy allows the query from your network. ☞ The snmpd daemon is running on the SNMP client. Troubleshooting to get functioning SNMP queries is important as many other supporting applications such as MRTG—which I’ll discuss next—rely on them in order to work correctly. ☞ You are querying using the correct SNMP version. ☞ Your /var/log/messages file does not contain errors that may have occurred while starting snmpd. .conf file is only read by the snmpd daemon when it starts up.mrtg.conf file so the configuration settings become active. MRTG MRTG (Multi-Router Traffic Grapher) is a public domain package for producing graphs of various router statistics via a Web page. A simple ping or traceroute test should be sufficient. ☞ You are using the correct community string. Simple SNMP Troubleshooting If your SNMP queries fail. ☞ Firewalls aren’t preventing SNMP queries from the SNMP client to the SNMP target. then verify that: ☞ You restarted your snmp.allow) and (/etc/hosts. ☞ Network routing between the client and target devices is correct.deny) should be blank.org) and also on your distribution CDs. Remember.

.” for installation details. When searching for the file. MRTG Download and Installation You need to install MRTG before proceeding.i386. Most Red Hat and Fedora Linux software products are available in the RPM format.10.2 A typical MRTG Web page.5-3. By default Apache expects the HTML files for your Web site to be located in /var/www/html. MRTG places its HTML files in /var/www/mrtg.rpm. you can refer to Chapter 20. you need to install the SNMP utility tools as explained earlier and you need to have a Web server package installed for MRTG to work.24Harrison_ch22. In addition to MRTG. remember that the MRTG RPM’s filename usually starts with mrtg and a version number. If you don’t get a positive response. The easiest way to tell if Apache is installed is to run the rpm -q httpd command. Red Hat Linux usually comes with the Apache Web server software preinstalled.qxd 2/25/05 10:14 AM Page 374 374 Monitoring Server Performance Chapter 22 Figure 22. “The Apache Web Server. as in mrtg-2.

qxd 2/25/05 10:14 AM Page 375 MRTG 375 Configuring MRTG By default. it creates a directory called /etc/mrtg in which all future configuration files are stored. . Place all data files in the directory /var/www/mrtg:. this causes errors when Fedora Core 1 polls another Fedora Core 1 server. the errors prevent the graphs from being created correctly. [root@bigboy tmp]# cfgmaker --output=/etc/mrtg/mrtg.” For now. follow these steps: 1. MRTG maps the inbound and outbound data throughput rates on the device it is polling. You can avoid this problem by removing the option to get graph titles using the much less recognizable MAC address of the interfaces. Use MRTG’s cfgmaker command to create a configuration file named mrtg..cfg \ --global “workdir: /var/www/mrtg” -ifref=ip \ --global ‘options[_]: growright.cfg configuration file for the server.24Harrison_ch22. --base: Walking ifAdminStatus --base: Walking ifOperStatus --base: Writing /etc/mrtg/mrtg. . Unfortunately. . When the MRTG RPM is installed. such as CPU and memory usage.cfg [root@bigboy tmp]# Tip Using the -ifref=ip cfgmaker command option instructs MRTG to use the interface’s IP address as the title for each graph.bits’ \ craz33guy@localhost --base: Get Device Info on craz33guy@localhost: --base: Vendor Id: --base: Populating confcache --snpo: confcache craz33guy@localhost: Descr lo --> 1 --snpo: confcache craz33guy@localhost: Descr wlan0 --> 2 . are discussed in Chapter 23. “Advanced MRTG for Linux.. I’ll stick with the default configuration.... Methods for specifying other OIDs. To create a replacement default /etc/mrtg/mrtg.cfg for the server (Bigboy) using a Read Only community string of craz33guy..

1.168. The explanation for changing the language character set for MRTG is given in the “Troubleshooting MRTG” section. . there are different versions of SNMP. Here is an example of cfgmaker using an SNMP version 2 query of a router with an IP address of 192.1. Use MRTG’s indexmaker command to create a Web index page using your new mrtg. edit this file.old updating log file [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg.1. The --snmp-options option’s five colons before the 2 are important: [root@bigboy tmp]# cfgmaker -output=/etc/mrtg/192.1. the first time it is run. Edit /etc/mrtg/mrtg.168.cfg [root@bigboy tmp]# 4. If your query doesn’t work.100 was invalid as well Rateup WARNING: /usr/bin/rateup Can’t remove localhost_192.log to localhost_192.3.1. which doesn’t pass any network traffic at all.old updating log file Rateup WARNING: /usr/bin/rateup Can’t rename localhost_192.1.0.1. 3.1.3 2.html \ /etc/mrtg/mrtg.24Harrison_ch22.100 Rateup WARNING: /usr/bin/rateup The backup log file for localhost_192. and the file that governs this is /etc/cron.cfg file as a guide.168. MRTG has no data files to move. For MRTG to work correctly.100.100.168.cfg Rateup WARNING: /usr/bin/rateup could not read the primary log file for localhost_192. replacing all occurrences of /usr/bin/mrtg with env LANG[eq]C /usr/bin/mrtg. You can specify MRTG’s SNMP query version with the --snmp-options cfgmaker option.168.cfg 5. You’ll get an error the two times as MRTG tries to move old data files.168. The MRTG Web GUI expects to find the index file in the default MRTG Web directory of /var/www/mrtg/.d/mrtg.100.qxd 2/25/05 10:14 AM Page 376 376 Monitoring Server Performance Chapter 22 Tip As explained in the SNMP section. MRTG is run every five minutes by default.168.168.cfg. Run MRTG using /etc/mrtg/mrtg. A certain candidate would be the virtual loopback interface Lo: (with the IP address of 127.100. and remove the sections related to interfaces you don’t need to monitor.1. check to make sure you are using the required version and then check other SNMP configuration parameters on the target device.3.168.old updating log file [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg.1). so the format of the command would be: [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index.cfg Rateup WARNING: /usr/bin/rateup Can’t remove localhost_192.cfg as your argument three times. [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg.0. and naturally.cfg \ -ifref=ip --global “workdir: /var/www/mrtg” \ --snmp-options=:::::2 craz33guy@192.

to regenerate your Web index page.d/mrtg. Create a new configuration file using the steps from the previous section. Some configuration may need to be done. and then restart the cron daemon to re-read the configuration: [root@bigboy tmp]# service crond restart You could also create a script with the /usr/bin/mrtg/etc/mrtg/ entries in it and make cron run it every five minutes.cfg 4. Add a new MRTG line in /etc/cron.cfg 3. Getting MRTG to Poll Multiple Devices The Fedora Core MRTG installation process creates a cron file named /etc/cron.1.0 network: .cfg.conf that includes all the necessary Apache commands for MRTG to work. This way you can just edit the script each time you add a device without having to restart cron. choose a filename that is not mrtg.cfg /etc/mrtg/device1.html \ /etc/mrtg/mrtg. With Fedora Core. You can add your home network to the file by inserting the network on the Allow from line. Other versions of Linux keep their MRTG cron entries inside the /etc/crontab file. This file tells the cron daemon to run MRTG using the /etc/mrtg/mrtg. Run the indexmaker command.24Harrison_ch22.cfg 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/device2.168.cfg 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/device1. You can configure MRTG to poll multiple devices. 2. but first I’ll show you how to poll multiple devices. Here’s how: 1.cfg /etc/mrtg/device2.d/mrtg for each new configuration file you create: 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/mrtg. Edit this file using the same syntax as the Fedora /etc/cron.cfg file every five minutes to poll your network devices. This will be covered later.d/mrtg file. or you can allow universal access by commenting out that line along with the Deny from line. MRTG creates an add-on configuration file named /etc/httpd/conf. each with a separate configuration file.qxd 2/25/05 10:14 AM Page 377 MRTG 377 This isn’t all—you need to view the graphs too. /etc/mrtg configu- [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index. and include all of your ration files. This example adds access from the 192.d/mrtg.cfg Configuring Apache to Work with MRTG MRTG is useful because it can provide a graphical representation of your server’s performance statistics via a Web browser. because by default MRTG accepts Web requests from the Linux console only. device.

Chapter 23 has detailed examples and explanations of how to monitor Linux disk. MRTG provides only network interface statistics. Note With newer versions of Fedora. With Fedora Core 1.htaccess file as described in Chapter 20. monthly. www.168. How to View the MRTG Graphs in Your Web Browser You can now access your MRTG graphs by pointing your browser to the URL: http://server-ip-address/mrtg/ Using MRTG to Monitor Other Subsystems MRTG will generate HTML pages with daily.conf to find it.24Harrison_ch22. you may want to add password protection to the directory by using a .0/24 </Location> If you want to access MRTG from the Internet.allow Allow from all </Location> Remember to restart Apache after you make these modifications for the changes to take effect.org. memory. then you’ll have to comment out the Deny statement and allow from all IP addresses: <Location /mrtg> Order deny. You can do this yourself by inserting this line at the very bottom of the main Apache configuration file before restarting Apache for the change to take effect: include “/etc/httpd/conf. . CPU. weekly. and yearly statistics for your interfaces.d/mrtg.d/ directory. and Web connection data.mrtg. also has links to other sites that show you how to monitor many other subsystems on a variety of devices and operating systems.1. you have to specifically configure the Apache configuration file /etc/httpd/conf/httpd. The MRTG Web site.conf” Basic Security If you are accessing MRTG graphs from the Internet.qxd 2/25/05 10:14 AM Page 378 378 Monitoring Server Performance Chapter 22 <Location /mrtg> Order deny. Apache automatically reads the add-on files in the /etc/httpd/conf.allow Deny from all Allow from localhost 192. By default.

other things may fail to work. and the current Fedora default of en_US.. When using a -ifref=ip statement with the cfgmaker command.24Harrison_ch22. which is the much harder to recognize interface MAC address. Basic Steps MRTG won’t work if SNMP queries don’t work. When it works. then you’ll get errors such as this every time you run MRTG from the command line or as part of a cron job: root@bigboy tmp]# mrtg /etc/mrtg/mrtg. where this is not the case: env LANG=C /usr/bin/mrtg ..qxd 2/25/05 10:14 AM Page 379 MRTG 379 Troubleshooting MRTG There are many simple steps you can use to troubleshoot MRTG. This is not necessarily a good idea. but en_US will after a system reboot. A better solution is to always run MRTG using this command instead of using just plain /usr/bin/mrtg: env LANG=C /usr/bin/mrtg This will modify the character set used by MRTG alone and shouldn’t affect anything else.. Upgrading to the most current Core 1 version of MRTG will fix the problem: .UTF-8 won’t work. Take a look at some of the most common ones. every line in the configuration file that is generated becomes commented out. If it isn’t set. ------------------------------------------------------------------[root@bigboy tmp]# Your system’s character set is defined in /etc/sysconfig/i18n. Setting the Correct Character Set MRTG usually works only if your system uses an ASCII-based (Western European) character set. this statement is very convenient. because it makes MRTG provide graphs sorted by the IP addresses of the interfaces instead of the default. especially if the native language Linux uses on your system is not ASCII based. Make sure you follow the SNMP troubleshooting steps if you have any difficulties.cfg ------------------------------------------------------------------ERROR: Mrtg will most likely not work propperly when the environment variable LANG is set to UTF-8. Fedora Core 1 MRTG Errors with net-snmp A bug appears in the MRTG implementation for some Fedora Core 1 MRTG versions when polling another Fedora Core 1 server. Please run mrtg in an envir.

1. Webalizer reads your Apache log files and creates a set of Web pages that enable you to view Web surfer statistics for your site.1.168.100:craz33guy@localhost: # SetEnv[localhost_192.6.0.100]: Traffic Analysis for 192.0.168.1.168.3.168.168.cfg Use of uninitialized value in hash element at /usr/bin/indexmaker line 307.100]: MRTG_INT_IP=”192.168.1].html file and gives errors: [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/stats/index.1.1.6) ### SNMPv1_Session (remote host: “localhost” [127.0 bits/s</TD></TR> # </TABLE> As all the lines in the configuration file are commented out with a # character.100 # PageTop[localhost_192.1.1.100]: /192.1.2.24Harrison_ch22. [root@bigboy tmp]# WEBALIZER Webalizer is a Web server log file analysis tool that comes installed by default on Red Hat/Fedora Linux. The information provided includes a list of your Web site’s most popular pages sorted by hits along with traffic graphs showing the times of day when your site is most popular.1.10.100’ ### ### The following interface is commented out because: ### * has a speed of which makes no sense ### * got ‘Received SNMP response with error code ### error status: noSuchName ### index 1 (OID: 1.100 # <TABLE> # <TR><TD>System:</TD> <TD>bigboy in Unknown</TD></TR> # <TR><TD>Maintainer:</TD> <TD>root@localhost</TD></TR> # <TR><TD>Description:</TD><TD> </TD></TR> # <TR><TD>ifType:</TD> <TD> ()</TD></TR> # <TR><TD>ifName:</TD> <TD></TD></TR> # <TR><TD>Max Speed:</TD> <TD>0.1.100” MRTG_INT_DES # MaxBytes[localhost_192.168.html /etc/mrtg/mrtg.168.168.2. indexmaker fails to create an index. .qxd 2/25/05 10:14 AM Page 380 380 Monitoring Server Performance Chapter 22 ### Interface 6 >> Descr: ‘’ | Name: ‘’ | Ip: ‘192.1.1. Each night.168.2.100]: 0 # Title[localhost_192.1.100]: <H1>Traffic Analysis for 192.161) ### community: “craz33guy” ### request ID: 824482716 ### PDU bufsize: 8000 bytes ### timeout: 2s ### retries: 5 # # Target[localhost_192.

but you may want to adjust the directory in which Webalizer places your graph statistics.0 0. The default settings should be sufficient for your Web server.0% hi.1 0:03. 4192844k free. be aware that running in quiet mode could hide deeper problems that could occur in the future.0 0:00. 95.qxd 2/25/05 10:14 AM Page 381 The top Command 381 How to View Your Webalizer Statistics izer. 868484k used.00 events/0 4 root 7 -10 0 0 0 S 0. As in the case of the MRTG add-on file mentioned above.0 0:00.85 init 2 root 34 19 0 0 0 S 0. 1 running.00. 276272k Swap: 4192924k total. 23:35.1% id.0 0.1% wa. however. 0. load average: 0. Webalizer places its index page in the directory /var/www/html/usage and allows you to view your data by visiting the URL http://server-ip-address/usage.0% ni. You also have to restart Apache to make the changes take effect. 2 users. 0. THE TOP COMMAND You can monitor the amount of memory and CPU resources your system is using with the top command: [root@bigboy tmp]# top top .5% sy.00 Tasks: 72 total.d/webalthat includes all the necessary Apache commands for Webalizer to work. 0 stopped.0 0:00.conf. 356256k PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND 1 root 16 0 2268 516 1420 S 0. This can be adjusted with the OutputDir directive in the file. you have to edit it to allow access to the Webalizer pages from locations other than the Linux console.0 0. 0.32 ksoftirqd/0 3 root 5 -10 0 0 0 S 0. Mem: 1027784k total. 71 sleeping. By default. 0 zombie Cpu(s): 4.conf Fedora creates an add-on configuration file named /etc/httpd/conf.3% us.0 0. 0. 159300k free.20:00:54 up 61 days. 80k used.24Harrison_ch22.02 khelper [root@bigboy tmp]# . 0. The Webalizer Configuration File Webalizer stores its configuration in the file /etc/webalizer. After adjustments. Webalizer functions with few annoyances.

memory is still 15% free (158. Here’s a sample: [root@bigboy tmp]# vmstat procs —————-memory————— —-swap— ——-io—— r b swpd free buff cache si 0 0 80 158196 276272 357236 0 [root@bigboy tmp]# so 0 bi 1 bo 0 . THE VMSTAT COMMAND You can also determine memory and swap usage with the vmstat command.--cpu-r b swpd free buff cache si so bi bo sy id wa 0 0 80 158196 276272 357236 0 0 1 0 1 95 0 [root@bigboy tmp]# in 0 cs us 4 4 As your memory fills up. In the example. Excess swapping of programs and data between disk and memory can cause your system to slow down significantly and memory usage should be monitored to allow you to plan ways to either increase RAM or tune the way your system operates. such as poor duplex negotiation. the server doesn’t seem to be swapping idle processes from memory to the swap disk partition as it isn’t being used at all. your system temporarily stores programs and data on your hard disk’s swap partition. THE FREE UTILITY The free utility can determine the amount of free RAM on your system. and network congestion due to excessive traffic.-system.1% idle and 15% of memory (159300k) is free. then start looking at networking problems.24Harrison_ch22. [root@bigboy tmp]# vmstat procs ------memory----. If your system seems slow but the CPU and memory usage is low. The amount of free memory may appear low.qxd 2/25/05 10:14 AM Page 382 382 Monitoring Server Performance Chapter 22 Here the CPU usage is running 95. but in this case. Excessive swapping can cause your system to slow down dramatically.780) and only 80MB of the swap partition is being used. the simplest ways to avoid this is to add more RAM or reduce the number of processes or users that are active on your system.---io-. bad cables.—-swap.196MB used from a total of 130. The output is easier to understand than vmstat’s. which provides a summary of what top produces.

because it can help you predict when things are going to go wrong or long-term trends in your Web traffic. consider running more efficient versions of programs. but also to graph many of the statistics listed in top. and vmstat. free. CONCLUSION Server monitoring is always a good practice. upgrading the capacity of your RAM. and. . Chapter 23 shows you how.24Harrison_ch22.qxd 2/25/05 10:14 AM Page 383 Conclusion 383 You should generally try to make your system run with at least 20% free memory on average. which should allow it to handle moderate spikes in usage caused by running memory-intensive cron batch jobs or tape backups. MRTG can be expanded not only to monitor traffic on your server’s NIC cards. of course. offloading some applications to servers with less load. If you cannot achieve this.

qxd 2/25/05 10:14 AM Page 384 .24Harrison_ch22.

All the chapter’s examples assume that the SNMP Read Only string is craz33guy and that the net-snmp-utils RPM package is installed (see Chapter 22. LOCATING AND VIEWING THE CONTENTS OF LINUX MIBS Residing in memory. 385 . The MIB configuration text files are located on your hard disk and loaded into memory each time SNMP restarts.25Harrison_ch23. You may also want to see graphs of CPU.qxd 2/25/05 10:15 AM Page 385 C H A P T E R 23 Advanced MRTG for Linux IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Locating and Viewing the Contents of Linux MIBs Testing Your MIB Value Differences in MIB and MRTG Terminology The CPU and the Memory Monitoring MIB The TCP/IP Monitoring MIB Manually Configuring Your MRTG File Implementing Advanced Server Monitoring Monitoring Non-Linux MIB Values Troubleshooting Conclusion In many cases using MRTG in a basic configuration to monitor the volume of network traffic to your server isn’t enough. This chapter explains how to find the values you want to monitor in the SNMP MIB files and then how to use this information to configure MRTG. and memory usage. disk. “Monitoring Server Performance”). MIBs are data structures that are constantly updated via the SNMP daemon.

.txt: “The number of times TCP connections have made a . In this snippet of the RFC1213-MIB... MIBs are very complicated. [root@silent mibs]# grep -i tcp /usr/share/snmp/mibs/*.qxd 2/25/05 10:15 AM Page 386 386 Advanced MRTG for Linux Chapter 23 You can easily find your Fedora Linux MIBs by using the locate command and filtering the output to include only values with the word “snmp” in them. RFC1213-MIB.txt . tcpActiveOpens OBJECT-TYPE SYNTAX Counter ACCESS read-only STATUS mandatory DESCRIPTION “The number of times TCP connections have made a direct transition to the SYN-SENT state from the . Don’t change them. such as the amount of packets passing through a NIC or amount of time the CPU has been busy since boot time. . MIBs usually track two types of values.txt: “The limit on the total number of TCP connections RFC1213-MIB. [root@bigboy tmp]# As the MIB configurations are text files you can search for keywords in them using the grep command. TCP-MIB.. you can see that querying the tcpActiveOpens object returns the number of active open TCP connections to the server. .txt file.. [root@silent mibs]# You can use the vi editor to look at the MIBs.. Integer values change instant by instant and are useful for tracking such statistics as the amount of memory currently being used.txt: “The number of times TCP connections have made a . because doing so could cause SNMP to fail.mib2c /usr/share/snmp/mibs /usr/share/snmp/mibs/DISMAN-SCHEDULE-MIB. the MIBs are located in the /usr/share/snmp/mibs directory: [root@bigboy tmp]# locate mib | grep snmp /usr/share/doc/net-snmp-5.0.. but fortunately the key sections are commented.txt \ | grep connections .6/README. Counter values are used for items that continuously increase as time passes. Each value tracked in a MIB is called an object and is often referred to by its object ID or OID..25Harrison_ch23. This example searches for the MIBs that keep track of TCP connections and returns the RFC1213 and TCP MIBs as the result.... As you can see in this case. The SYNTAX field shows that this is a counter value..

MRTG considers all values to be counters. . The snmpget command doesn’t work with branches. snmpget are the ones you DIFFERENCES IN MIB AND MRTG TERMINOLOGY Always keep in mind that MRTG refers to MIB counter values as counter values. it plots only how much the value has changed since the last polling cycle. Follow up with the snmpwalk command shows that the value is located in ssCpuRawUser.” ::= { tcp 5 } You’ll explore the differences between SNMP and MRTG terminologies in more detail later. the ssCpuRawUser OID alias was found to be interesting. MRTG doesn’t plot counter values as a constantly increasing graph. fortunately. In the example below.1.0 instead. [root@bigboy tmp]# snmpget -v1 -c craz33guy localhost ssCpuRawUser Error in packet Reason: (noSuchName) There is no such variable name in this MIB. By default. giving an error stating that the MIB variable couldn’t be found. but the snmpget command fails to get a value. Many times the text aliases in an MIB reference the OID branch only and not the data located in a leaf ending in an additional number.25Harrison_ch23. Failed object: UCD-SNMP-MIB::ssCpuRawUser [root@bigboy tmp]# [root@bigboy tmp]# snmpwalk -v1 -c craz33guy localhost ssCpuRawUser UCD-SNMP-MIB::ssCpuRawUser.qxd 2/25/05 10:15 AM Page 387 Differences in MIB and MRTG Terminology 387 CLOSED state.0 UCD-SNMP-MIB::ssCpuRawUser. you can edit your MRTG configuration file to make it graph this information in a percentage use format (more on this later). The snmpget is then successful in retrieving the counter32 type data with a current value of 396271. you can use the snmpwalk command to poll it. CPU usage is typically tracked by MIBs as a counter value. TESTING YOUR MIB VALUE Once you identify an interesting MIB value for your Linux system. Understanding them will be important in understanding how to use MRTG to track MIB values.0 or . such as .0 = Counter32: 396271 [root@bigboy tmp]# The MIB values that work successfully with should use with MRTG. It refers to MIB integer and gauge values as gauge.0 = Counter32: 396241 [root@bigboy tmp]# snmpget -v1 -c craz33guy localhost ssCpuRawUser.

THE CPU AND THE MEMORY MONITORING MIB The UCD-SNMP-MIB MIB keeps track of a number of key performance MIB objects.1 Important Objects in the UCD-SNMP-MIB MIB MIB Type Counter MRTG Type Counter Description Total CPU usage by applications run by nonprivileged users since the system booted. including the commonly used ones in Table 23. I’ll come back to these later. The percentage of the time the CPU is running idle.1. the MIB object name. Table 23. Subtracting this value from 100 can give a good approximation of total CPU usage.qxd 2/25/05 10:15 AM Page 388 388 Advanced MRTG for Linux Chapter 23 The syntax type. Total CPU usage by applications run by privileged system processes since the system booted. system.25Harrison_ch23. and the description of what it does are the most important things you need to know when configuring MRTG. and nice values can give a good approximation of total CPU usage. Adding the user.2 details the most important objects in TCP-MIB. Available physical memory space on the host. UCD-SNMP-MIB Object Variable ssCpuRawUser ssCpuRawSystem Counter Counter ssCpuRawNice Counter Counter ssCpuRawIdle Counter Counter memAvailReal Integer Gauge THE TCP/IP MONITORING MIB The TCP-MIB MIB keeps track of data connection information and contains the very useful tcpActiveOpens and tcpCurrEstab objects. . Total CPU usage by applications running at a nondefault priority level. Table 23.

Parameter Formats MRTG configuration parameters are always followed by a graph name surrounded by square brackets and a colon.2 Variable Important Objects in the TCP-MIB MIB MIB Type Counter Gauge Counter MRTG Type Counter Gauge Counter Description Measures the number of completed TCP connections. which is usually placed at the very top of the graph name list.25Harrison_ch23. the parameters for a particular graph are usually grouped together. . others refer to the Summary View. The two OID values are separated by an & symbol. and annual graphs. The mrtg and indexmaker program can be fed individual OIDs from a customized configuration file and will function as expected if you edit the file correctly. If you want to monitor any other pairs of OIDs.qxd 2/25/05 10:15 AM Page 389 Manually Configuring Your MRTG File 389 Table 23. Measures the number of TCP connections in the established state. Some of the parameters in the configuration file refer to the Detailed View. Total number of TCP segments with bad checksum errors. You have to click on the Summary View graphs to get the Detailed View page behind it with the daily. The format looks like this: Parameter[graph name]: value For ease of editing. and the second one is the output OID. simultaneously tracking two OIDs: the NIC’s input and output data statistics. The indexmaker program also uses this information to create the overview. Each graph can track two OIDS listed in the Target parameter. monthly. you have to manually create the configuration files. weekly. The mrtg program then uses these configuration files to determine the type of data to record in its data directory. because cfgmaker isn’t aware of any OIDs other than those related to a NIC. or Summary View Web page for the MIB OIDs you’re monitoring. This Summary View page shows daily statistics only. the first one is the input OID. TCP-MIB Object tcpActiveOpens tcpCurrEstab tcpInErrs MANUALLY CONFIGURING YOUR MRTG FILE The MRTG cfgmaker program creates configuration files for network interfaces only.

The growright option makes sure the data at the right of the screen is for the most current graph values. This usually makes the graphs more intuitively easy to read. Options Parameters Options parameters provide MRTG with graph formatting information. Likewise. The Ylegend is the legend for the Y axis. In the case of a default MRTG configuration this would be the data flow through the interface in bits or bytes per second. and current values of the graph’s OID statistics. If the value you are monitoring is in bytes. The gauge option alerts MRTG to the fact that the graphed values are of the gauge type. Here are some examples for two different graphs: options[graph1]: growright. then you can convert the output to bits using the bits option. Here is an example of the legends of a default MRTG configuration: YLegend[graph1]: Legend1[graph1]: Legend2[graph1]: LegendI[graph1]: LegendO[graph1]: Bits per second Incoming Traffic in Bits per Second Outgoing Traffic in Bits per Second In Out You can prevent MRTG from printing the legend at the bottom of the graph by leaving the value of the legend blank like this: LegendI[graph1]: Later you’ll learn how to match the legends to the OIDs for a variety of situations.nopercent. You can use the legendI parameter for the description of the input graph (first graph OID) and the legendO for the output graph (second graph OID).25Harrison_ch23.bits . The nopercent option prevents MRTG from printing percentage style statistics in the legends at the bottom of the graph. the value you are trying to compare. you can convert per second values to per minute graphs using the perminute option. each graph has a legend that shows the max. The space available under each graph’s legend is tiny so MRTG also has legend1 and legend2 parameters that are placed at the very bottom of the page to provide more details. average. Parameter legend1 is the expansion of legendI. and legend2 is the expansion of legendO. MRTG defaults to growing from the left.perminute options[graph2]: gauge.qxd 2/25/05 10:15 AM Page 390 390 Advanced MRTG for Linux Chapter 23 Legend Parameters On the Detailed View Web page.

you may want to turn off MRTG’s auto scaling. and daily views on the Detailed View page and gives them a maximum value of 100: Unscaled[graph1]: ymwd MaxBytes[graph1]: 100 Defining The MIB Target Parameters As stated before. with scaling. weekly. You have to specify the two MIB OID objects—the SNMP password and the IP address of the device you are querying in this parameter—and separate them with an & character: Target[graph1]: mib-object-1.qxd 2/25/05 10:15 AM Page 391 Manually Configuring Your MRTG File 391 If you place this parameter at the top with a label of [_] it gets applied to all the graphs defined in the file: options[_]: growright Title Parameters The title on the Summary Page is provided by the Title parameter. so that the vertical peak is near the top of the graph image. This is so even if you set the MaxBytes parameter.0:<SNMP-password>@<IP-address> .0&mib-object-2. MRTG also tries to adjust its graphs so that the largest value plotted on the graph is always close to the top.25Harrison_ch23. MRTG will have a vertical plot of 0% to 60%. If you are plotting percentage CPU usage. the PageTop parameter tells the title for the Detailed View page. then the peak will be only 60% of the way up as the graph plots from 0% to 100%. When you are plotting a value that has a known maximum and you always want to have this value at the top of the vertical legend. and MaxBytes is set to 100. When scaling is off. and the server reaches a maximum of 60%. The PageTop string must start with <H1> and end with </H1>. Title[graph1]: Interface eth0 PageTop[graph1]: <H1>Detailed Statistics For Interface eth0</H1> Scaling Parameters The MaxBytes parameter is the maximum amount of data MRTG will plot on a graph. Anything more than this seems to disappear over the edge of the graph. monthly. MRTG always tries to compare two MIB OID values that are defined by the Target parameter. The example removes scaling from the yearly.

3 Legend Legend1 Legend2 LegendI LegendO Mapping MIBs to the Graph Legends Maps to Target MIB 1 2 1 2 So in the example below.qxd 2/25/05 10:15 AM Page 392 392 Advanced MRTG for Linux Chapter 23 The numeric value.0 is tagged onto the end of the output.25Harrison_ch23.3. in this case . which plots only mib-object-1.0 and mib-object-4.0&mib-object-2. Target[graph1]: mib-object-1. Target[graph1]: mib-object-1.0 for the other. Notice how the .0&mib-object-4.0&mib-object-2. legend1 and legendI describe and legend2 and legendO describe mib-object-2.0&mib-object-1. mib-object-1.0 and mib-object-3.0:<SNMP-password>@<IPaddress> + mib-object-3. Table 23. [root@silent mibs]# snmpwalk -v 1 -c craz33guy localhost ssCpuRawUser UCD-SNMP-MIB::ssCpuRawUser. The next example adds mib-object-1.0:<SNMP-password>@<IPaddress> Adding MIB Values Together for a Graph You can use the plus sign between the pairs of MIB object values to add them together. at the end of the MIB is required. The next example uses the SNMP command to return the user mode CPU utilization of a Linux server. The resulting MRTG graph actually superimposes the input and output graphs one on top of the other.0.0 = Counter32: 926739 [root@silent mibs]# The MRTG legends map to the MIBs listed in the target as shown in Table 23.0.0:<SNMP-password>@<IPaddress> Plotting Only One MIB Value If you want to plot only one MIB value.0:<SNMP-password>@<IP-address> . you can just repeat the target MIB in the definition as in the next example. Target[graph1]: mib-object-1.0.0 for one graph and adds mib-object-2.

0& tcpPassiveOpens. and a few processes running in nice mode. There must be white spaces before and after all these operators for MRTG to work correctly. Notice how the mandatory white spaces separate the mathematical operators from the next target element.0:<SNMP-password>@<IPaddress> + ssCpuRawSystem. and divide (%). If not.0&ssCpuRawUser.0:<SNMPpassword>@<IP-address> MaxBytes[graph1]: 1000000 Options[graph1]: perminute .0&ssCpuRawSystem. The tcpActiveOpens MIB object monitors new connections originating from the server.0:<SNMP-password>@<IP-address> Be sure to place this command on a single line.0&ssCpuRawNice.qxd 2/25/05 10:15 AM Page 393 Manually Configuring Your MRTG File 393 You can use other mathematical operators.growright.growright. Sample Target: Total CPU Usage Linux CPU usage is occupied by system processes.0&memTotalReal.0:<SNMP-password>@<IPAddress> ) * 100 / ( memTotalReal. Target[graph1]: ( memAvailReal. Left and right parentheses are also valid.0:<SNMPpassword>@<IP-Address> ) options[graph1]: nopercent. This example adds them all together in a single plot: Target[graph1]:ssCpuRawUser. Notice that this is a gauge type variable: Target[graph1]: memAvailReal.0& memAvailReal.25Harrison_ch23.gauge Sample Target: Newly Created Connections HTTP traffic caused by Web browsing usually consists of many very shortlived connections.0:<SNMP-password>@<IPaddress> + ssCpuRawNice. plot the percentage of available memory. such as subtract (-). user mode processes. multiply (*).0:<SNMP-password>@<IPaddress> options[graph1]: nopercent. you’ll get oddly shaded graphs. On smaller Web sites you may want to use the perminute option to make the graphs more meaningful: Target[graph1]: tcpPassiveOpens. The tcpPassiveOpens MIB object tracks newly created connections and is suited for this type of data transfer.0&memTotalReal.gauge Next. Sample Target: Memory Usage Here is an example for plotting the amount of free memory versus the total RAM installed in the server.

Use the snmpwalk command to query the the dskPercent dskPercent. Add two entries to your snmpd.1 UCD-SNMP-MIB::dskPercent.25Harrison_ch23.1& dskPercent. Object (/home).0:<SNMP-password>@<IPaddress> MaxBytes[graph1]: 1000000 Options[graph1]: gauge Sample Target: Disk Partition Usage In this example. such as FTP and SSH. create longer established connections while people download large files or stay logged into the server.conf file: disk disk /home /var 3. The tcpCurrEstab MIB object measures the total number of connections in the established state and is a gauge value: Target[graph1]: tcpCurrEstab.1 refers to the first disk entry in snmpd.qxd 2/25/05 10:15 AM Page 394 394 Advanced MRTG for Linux Chapter 23 Sample Target: Total TCP Established Connections Other protocols.2 refers to the second (/var). Restart the SNMP daemon to reload the values: [root@bigboy tmp]# service snmpd restart 4.gauge .1 = INTEGER: 13 [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost \ dskPercent.1:<SNMP-password>@<IPaddress> options[graph1]: growright. you’ll monitor the /var and /home disk partitions on the system: 1.2 = INTEGER: 73 [root@bigboy tmp]# Your MRTG target for these gauge MIB objects should look like this: Target[graph1]: dskPercent. First use the df -k command to get a list of the partitions in use: Used Available Use% Mounted 128199 19178 122864 8229 890092 171832 351302 76692 860188 471272 2832140 1261520 27% 21% 13% 2% 24% 73% / /boot /home /tmp /usr /var [root@bigboy tmp]# df -k Filesystem 1K-blocks on /dev/hda8 505605 /dev/hda1 101089 /dev/hda5 1035660 /dev/hda6 505605 /dev/hda3 3921436 /dev/hda2 1510060 [root@bigboy tmp]# 2. MIB.0&tcpCurrEstab.conf dskPercent.2 UCD-SNMP-MIB::dskPercent. and [root@bigboy tmp]# snmpwalk -v 1 -c craz33guy localhost \ dskPercent.

qxd 2/25/05 10:15 AM Page 395 Implementing Advanced Server Monitoring 395 Defining Global Variables You have to make sure MRTG knows where the MIBs you’re using are located. /usr/share/snmp/mibs/TCP-MIB. and then you can integrate it into the existing MRTG configuration. memory. Specify their locations with the global LoadMIBs parameter.cpu]: <H1>CPU Load .cfg # # Configuration file for non bandwidth server statistics # localhost for # # Define global options # LoadMIBs: /usr/share/snmp/mibs/UCD-SNMPMIB. You must also define where the HTML files will be located. disk.0&ssCpuRawSystem.txt workdir: /var/www/mrtg/ IMPLEMENTING ADVANCED SERVER MONITORING You now can combine all you have learned to create a configuration file that monitors all these variables.cpu]: Server CPU Load PageTop[server.0&ssCpuRawUser.System.cpu]:ssCpuRawUser.txt. A Complete Sample Configuration Here is a sample configuration file that is used to query server CPU. and TCP connection information: # # File: /etc/mrtg/server-info. The default location MRTG uses may not be valid.0:craz33guy@localhost + ssCpuRawNice. the example specifies the default Fedora MRTG HTML directory: LoadMIBs: /usr/share/snmp/mibs/UCD-SNMP-MIB.0&ssCpuRawNice./usr/share/snmp/mibs/TCP-MIB.txt.0:craz33guy@localhost + ssCpuRawSystem.txt workdir: /var/www/mrtg/ # # CPU Monitoring # (Scaled so that the sum of all three values doesn’t exceed 100) # Target[server.25Harrison_ch23.0:craz33guy@localhost Title[server. User and Nice Processes</H1> .

cpu]: Options[server.nopercent kMG[server.memory]: Total Legend1[server.cpu]: 100 ShortLegend[server.newconns]: c/s .nopercent Unscaled[server.memory]: B YLegend[server. not including swap.T. in bytes Legend2[server.nopercent Unscaled[server.memory]: Bytes LegendI[server.newconns]: tcpPassiveOpens.mempercent]: Percentage Free Memory Legend2[server.cpu]: CPU Utilization Legend1[server.newconns]: <H1>New TCP Connections</H1> MaxBytes[server.mempercent]: Percentage Free Memory Target[server.gauge.mempercent]: <H1>Percentage Free Memory</H1> growright.memory]: Total memory Options[server.cpu]: % YLegend[server.memory]: memAvailReal.0:craz33guy@localhost ) * 100 / ( memTotalReal.mempercent]: Percentage Free Memory # # New TCP Connection Monitoring (per minute) # Target[server.X # # Memory Monitoring (Percentage usage) # Title[server.G.memory]: <H1>Free Memory</H1> MaxBytes[server.mempercent]: Free Legend1[server.transparent.0&memTotalReal.mempercent]: Percent LegendI[server.growright.cpu]: Used LegendO[server.25Harrison_ch23.memory]: k.mempercent]: ymwd MaxBytes[server.cpu]: ymwd # # Memory Monitoring (Total Versus Available Memory) # Target[server.0&tcpActiveOpens.mempercent]: Memory % ShortLegend[server.cpu]: growright.mempercent]: PageTop[server.0:craz33guy@localhost Title[server.memory]: gauge.memory]: Free Memory PageTop[server.mempercent]: ( memAvailReal.qxd 2/25/05 10:15 AM Page 396 396 Advanced MRTG for Linux Chapter 23 MaxBytes[server.memory]: Free memory.cpu]: Current CPU percentage load LegendI[server.0:craz33guy@localhost Title[server.P.0&memTotalReal.M.0:craz33guy@localhost ) options[server.mempercent]: Free LegendO[server.mempercent]: 100 YLegend[server.newconns]: 10000000000 ShortLegend[server.memory]: 100000000000 ShortLegend[server.memory]: Free LegendO[server.0&memAvailReal.newconns]: Newly Created TCP Connections PageTop[server.

MRTG then creates the files.estabcons]: <H1>Established TCP Connections</H1> MaxBytes[server.cfg .2:craz33guy@localhost Title[server. Run the /usr/bin/mrtg command followed by the name of the configuration file three times.1&dskPercent. Restart SNMP to make sure the disk monitoring commands in the snmpd.estabcons]: Established connections Legend2[server.newconns]: Conns / Min In Out New inbound connections New outbound connections growright.gauge # # Disk Usage Monitoring # Target[server.disk]: Utilization LegendI[server.estabcons]: In LegendO[server.25Harrison_ch23.0:craz33guy@localhost Title[server.nopercent. If all goes well.estabcons]: 10000000000 ShortLegend[server.disk]: /var Options[server.newconns]: Options[server. all the files are created and MRTG should operate smoothly. By the third run.disk]: ymwd Testing the Configuration The next step is to test that MRTG can load the configuration file correctly.disk]: gauge. [root@bigboy tmp]# service snmpd restart [root@bigboy tmp]# env LANG=C /usr/bin/mrtg /etc/mrtg/server-stats.perminute # # Established TCP Connections # Target[server.disk]: 100 ShortLegend[server.disk]: /home LegendO[server.newconns]: Legend2[server.conf file are activated.estabcons]: Currently Established TCP Connections PageTop[server.estabcons]: Legend1[server.estabcons]: tcpCurrEstab.estabcons]: growright.disk]: Disk Partition Usage PageTop[server.qxd 2/25/05 10:15 AM Page 397 Implementing Advanced Server Monitoring 397 YLegend[server.estabcons]: YLegend[server.nopercent.newconns]: LegendO[server.nopercent Unscaled[server.growright.disk]: % YLegend[server.disk]: dskPercent.0&tcpCurrEstab.newconns]: Legend1[server.estabcons]: Options[server.disk]: <H1>Disk Partition Usage /home and /var</H1> MaxBytes[server. MRTG will complain only about the fact that certain database files don’t exist.estabcons]: Connections LegendI[server.newconns]: LegendI[server.

You will also have to restart cron with the service crond restart for it to read its new configuration file that tells it to additionally run MRTG every five minutes using the new MRTG configuration file. .cfg) plus the new one you created (/etc/mrtg/server-stats.cfg Some versions of Linux require you to edit your /etc/crontab file instead. You do not know what to do next. Downloading MIBs from Cisco and using them with the snmpget command was not a success.html \ /etc/mrtg/mrtg. other types of systems will need additional MIBs whose correct installation may be unclear in user guides or just not available. The basic MRTG configuration shown in Chapter 22 provides the data bandwidth statistics. In such cases. as well.qxd 2/25/05 10:15 AM Page 398 398 Advanced MRTG for Linux Chapter 23 Creating a New MRTG Index Page to Include this File Use the indexmaker command and include your original MRTG configuration file from Chapter 22 (/etc/mrtg/mrtg.25Harrison_ch23.d/mrtg file: 0-59/5 * * * * root env LANG=C /usr/bin/mrtg /etc/mrtg/serverstats.cfg /etc/mrtg/server-stats. Scenario Imagine that your small company has purchased a second-hand Cisco switch to connect its Web site servers to the Internet. [root@bigboy tmp]# indexmaker --output=/var/www/mrtg/index. add this line to your /etc/cron. you’ll need to know the exact value of the OID. To do so. See Chapter 22 for more details.cfg).cfg Configuring CRON to Use the New MRTG File The final step is to make sure that MRTG is configured to poll your server every five minutes using this new configuration file. [root@bigboy tmp]# service crond restart MONITORING NON-LINUX MIB VALUES All the MIBs mentioned so far are for Linux systems. but you want to measure the CPU load the traffic is having on the device.

1.1. Like most network equipment manufacturers.9.1.3.1.1.6.1. and 1.3. 1.1.1.1.1.6” “1.6. The SNMP files for Cisco’s devices can be found at ftp.109.1.1.1. so it is best to use the snmpwalk command to try to get all the values below this root first: [root@bigboy tmp]# snmpwalk -v1 -c craz33guy cisco-switch 1.9.5” “1. all the OIDs are a part of the same tree starting with 1.1.9.4.1.1.1.oid will contain the necessary values and find these entries inside it: “cpmCPUTotalPhysicalIndex” “cpmCPUTotal5sec” “cpmCPUTotal1min” “cpmCPUTotal5min” “cpmCPUTotal5secRev” “cpmCPUTotal1minRev” “cpmCPUTotal5minRev” “1.4. you can use this for your MRTG queries. OID value.1) in the original OID file maps to the word “enterprise”.5.1.109.1.1.3.3.3.4.1.1.1 = Gauge32: 32 SNMPv2-SMI::enterprises.1.1.109.1.1.1.1.4” “1.4.9. After looking at all the OID files.1.1.1.1.109.9.1.6.9.1.109.1.1.109.109.1.4.1.1.4.9.1.1.109.6.109.1.9.8” Testing The OIDs As you can see.4.2.9.6. OIDs are in the oid directory beneath that.3. you decide that the file CISCO-PROCESSMIB.1.6.1.9.3.1.9.9.25Harrison_ch23.1.5.1.9. 1.1.9.1.1.1.6.1.1.2” “1.9.1.109.1.1.109.4.1.1.3.9.1.1.1.1.qxd 2/25/05 10:15 AM Page 399 Monitoring Non-Linux MIB Values 399 Find the OIDs When MIB values fail.1.4.1.8 are not supported.6.9.1 = Gauge32: 33 [root@bigboy tmp]# Success! Now prises. Notice also how SNMP has determined that the first part of the OID value (1.1 SNMPv2-SMI::enterprises.9.1.4.1.9. it is best to try to find the exact OID value.9.109.9.6.109.9.5.cisco.1.9.1.1.1.1.1 = INTEGER: 0 SNMPv2-SMI::enterprises.9.3.1.109.9.1.109.6.9.1.1 = Gauge32: 32 [root@bigboy tmp]# Although listed in the OID file. Cisco has an FTP site from which you can download both MIBs and OIDs.1.1.9.1.6.9.1.9.1.1.1.9. The OIDs provided may be incomplete.4.109..1.6.1 SNMPv2-SMI::enterprises.7” “1.1.4. you can use the snmpget command to set only one of the OID values returned by snmpwalk: [root@bigboy tmp]# snmpwalk -v1 -c craz33guy cisco-switch 1.1.1.1.3.com in the /pub/mibs directory.3.1 = Gauge32: 32 SNMPv2-SMI::enterprises.1.1.1.1.1.1.3.3” “1. enter- .7.1.9. Next.9.

qxd 2/25/05 10:15 AM Page 400 400 Advanced MRTG for Linux Chapter 23 TROUBLESHOOTING The troubleshooting techniques for advanced MRTG are similar to those mentioned in Chapter 22. CONCLUSION Using the guidelines in this chapter you should be able to graph most SNMP MIB values available on any type of device. Verify the IP address and community string of the target device you intend to poll. 4. Check your MRTG parameters to make sure they are correct. Graph results that are eight times what you expect may have the bits parameter set. If not. Make sure you can do an SNMP walk of the target device. Gauge values defined as counter and vice versa will cause your graphs to have continuous zero values. flexible monitoring tool and should be considered as a part of any systems administrator’s server management plans. Ensure you can do an SNMP get of the specific OID value listed in your MRTG configuration file. revise your access controls on the target device and any firewall rules that may impede SNMP traffic. These quick steps should be sufficient in most cases and will reward you with a more manageable network. but because you have done some customizations you’ll have to go the extra mile: 1.25Harrison_ch23. . 3. 2. MRTG is an excellent.

you’ll need a central server that all can access.26Harrison_ch24. not all your servers will have Internet access. consult www.ntp. It also reduces Internet bandwidth usage due to NTP traffic and reduces the need to manage firewall rules for each NTP client on your network. This makes the correlation of system events on different systems much easier. in such cases.conf File How to Get NTP Started Testing and Troubleshooting NTP Configuring Cisco Devices to Use an NTP Server Firewalls and NTP Configuring a Windows NTP Client Conclusion The Network Time Protocol (NTP) is a protocol used to help synchronize your Linux system’s clock with an accurate time source. There are a number of sites that allow the general public to synchronize with them. Sometimes. They are divided into two types: ☞ Stratum 1: NTP sites using an atomic clock for timing ☞ Stratum 2: NTP sites with slightly less accurate time sources It is good practice to have at least one server on your network be the local time server for all your other devices. For a list of available Stratum 1 and 2 servers.qxd 2/25/05 10:16 AM Page 401 C H A P T E R 24 The NTP Server IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Download and Install the NTP Package The /etc/ntp.org. 401 .

1 . You do so with a modified restrict statement with the noquery replaced with a notrust keyword.255 nomodify The mask 255. Downloading and installing RPMs isn’t hard.gov 2. then you’ll have to define the networks from which this server will accept NTP synchronization requests.255.server.255.1.255. remember that the NTP RPM’s filename usually starts with ntp followed by a version number.rpm. This allows the network to query your NTP server. 3.0 notrust nomodify notrap In this case the mask statement has been expanded to include all 255 possible IP addresses on the local network.qxd 2/25/05 10:16 AM Page 402 402 The NTP Server Chapter 24 DOWNLOAD AND INSTALL THE NTP PACKAGE Most Red Hat and Fedora Linux software products are available in the RPM format. THE /ETC/NTP.org ntp. Make sure that localhost (the universal IP address used to refer to a Linux server itself) has full access without any restricting keywords: restrict 127. If this server is also going to provide time for other computers.i386. 4. such as PCs. The syntax is: restrict 192. “Installing RPM Software.2-5.research. but it won’t be trusted to be a source of NTP synchronization data.255. Restrict the type of access you allow these servers. other Linux servers.research.255 statement is really a subnet mask limiting access to the single IP address of the remote NTP servers. Specify the servers you’re interested in: server server otherntp.org # A stratum 2 server at research.26Harrison_ch24.255. and networking devices. as in ntp-4.” has all the details.conf file is the main configuration file for Linux NTP in which you place the IP addresses of the Stratum 1 and Stratum 2 servers you want to use.168. Here are the steps to create a configuration file using a pair of sample Internet-based NTP servers: 1.255. In this example. When searching for the file.1.255. the servers are not allowed to modify the run-time configuration or query your Linux NTP server restrict otherntp. If you need a refresher.gov notrap noquery mask 255. Chapter 6.org notrap noquery restrict ntp.CONF FILE The /etc/ntp.server.255 nomodify mask 255.gov # A stratum 1 server at server.0.0 mask 255.255.0.

The ntpdate command doesn’t run continuously in the background. your NTP daemon will eventually terminate itself leaving an error message in the /var/log/messages file. Here are some guidelines you can follow to get NTP working correctly. Verifying NTP is Running To test whether the NTP process is running. ☞ To get NTP configured to start at boot. and restart NTP after booting.qxd 2/25/05 10:16 AM Page 403 Testing and Troubleshooting NTP 403 5. HOW TO GET NTP STARTED You have to restart the NTP process every time you make a change to the configuration file for the changes to take effect on the running process. . was correctly set to 8:03 a. Take a look at some sample output of the ntpdate command in which a server whose initial time was set to midnight.m. you should test it to make sure it is working. Run the ntpdate -u command to force your server to become instantly synchronized with its NTP servers before starting the NTP daemon for the first time.26Harrison_ch24. stop. use the command: [root@bigboy tmp]# pgrep ntpd You should get a response of plain old process ID numbers. Save the file and restart NTP for the settings to take effect. you still have to run the ntpd daemon to get continuous NTP updates. Doing an Initial Synchronization If the time on the local server is very different from that of its primary time server. follow these examples: [root@bigboy tmp]# service ntpd start [root@bigboy tmp]# service ntpd stop [root@bigboy tmp]# service ntpd restart TESTING AND TROUBLESHOOTING NTP After configuring and starting NTP. use the line: [root@bigboy tmp]# chkconfig ntpd on ☞ To start. You can now configure other Linux hosts on your network to synchronize with this new master NTP server in a similar fashion.

100 offset 28993.168.100 offset 2. [root@bigboy tmp]# ntpq -p Here is some sample output of the command: . It provided you with a list of configured time servers and the delay.1.1.168.100 Looking for host 192.1.1. run ntpdate and then start ntpd again: [root@smallfry tmp]# service ntpd stop [root@smallfry tmp]# ntpdate -u 192.168.084943 sec [root@smallfry tmp]# ntpdate -u 192. but it must be run while the ntpd process is stopped.100 Looking for host 192.my-web-site.100.168.100 and service ntp host found : bigboy.qxd 2/25/05 10:16 AM Page 404 404 The NTP Server Chapter 24 ☞ The date was originally set to midnight which was verified by using the date command: [root@smallfry tmp]# date Thu Aug 12 00:00:00 PDT 2004 [root@smallfry tmp]# ☞ The ntpdate command is run three times to synchronize smallfry’s clock to server 192.org 12 Aug 08:03:42 ntpdate[2472]: step time server 192.168. So you’ll have to stop ntpd. the delay and offset values should be non-zero and the jitter value should be under 100.my-web-site.100 Looking for host 192.org 12 Aug 08:03:38 ntpdate[2472]: step time server 192. and jitter that your server is experiencing with them. For correct synchronization.1.1.084943 sec [root@smallfry tmp]# service ntpd start [root@smallfry tmp]# ☞ The date is now corrected: [root@smallfry tmp]# date Thu Aug 12 08:03:45 PDT 2004 [root@smallfry tmp]# Determining If NTP Is Synchronized Properly Use the ntpq command to see the servers with which you are synchronized.1.1.100 and service ntp host found : bigboy. offset.168.168.1.26Harrison_ch24.100 offset 0.168.org 12 Aug 08:03:40 ntpdate[2472]: step time server 192.467652 sec [root@smallfry tmp]# ntpdate -u 192.1.100 and service ntp host found : bigboy.my-web-site.168.168.

0.c 0.0 16 u 64 0 0. This could be located on one of the networks between the NTP server and its time source.strath harris.424 -3.net .anl.572 milo. .0.047 -ntp0. such as iptables. 2 u 972 1024 377 38.wisc.000 0.028 -dr-zaius.333 +taylor.aol-ca.993 -15. 1 u 426 1024 377 107.379 0. ☞ A firewall blocking access to your Stratum 1 and 2 NTP servers.127 3.edu 2 u 454 1024 347 54. In some versions of the Fedora Core 2’s implementation of NTP.go 2 u 818 1024 125 41.000 0.wisc.000 4000.cc.00 nist1.strat 3 u 507 1024 377 115.cs.mcs.000 0.usno.0. remote refid st t when poll reach delay offset jitter ====================================================================== ======= LOCAL(0) LOCAL(0) 10 l 64 7 0.cs.000 0.264 1.0 16 u .via. clients will not be able to synchronize with a Fedora Core 2 time server unless the notrust nomodify notrap keywords are removed from the NTP client’s restrict statement.681 -18.umb.392 -mailer1. or firewall software.549 1.025 1.000 4000.cis. ☞ The notrust nomodify notrap keywords are present in the restrict statement for the NTP client.0.psc. ben.000 0.wis ben.edu ntp1.1024 0 0.00 snvl-smtp1.26Harrison_ch24.0.0.0.sigmaso 3 u 95 1024 377 31.018 2. could be running on the server itself.qxd 2/25/05 10:16 AM Page 405 Testing and Troubleshooting NTP 405 remote refid st t when poll reach delay offset jitter ====================================================================== ======== -jj. In these cases.0 16 u 64 0 0.00 Your Linux NTP Clients Cannot Synchronize Properly A telltale sign that you haven’t got proper synchronization is when all the remote servers have jitters of 4000 with delay and reach values of 0.642 *clock.GPS.conectiv.edu gandalf.wisc.534 ntp1.008 ntp-cup.mcs.externa 0.cs.000 4000.098 3.go ntp0.00 This could be caused by: ☞ Older versions of the NTP package that don’t work correctly if you use the DNS name for the NTP servers.0 16 u 64 0 0.979 0.000 0.000 4000.589 28.edu 2 u 502 1024 357 55.cs.anl.206 19.cs.tr 0.274 -5.0. you use the actual IP addresses instead.trim 0.navy.

0 mask 255.1.168.0 mask 255.201.1. I have included the necessary NTP commands for a variety of Cisco Systems products. use the commands: ciscorouter> enable password: ********* ciscorouter# config t ciscorouter(config)# ntp update-calendar ciscorouter(config)# ntp server 192.255. in Fedora Core 2. Instead.168.255.26Harrison_ch24. the permissions on this directory don’t allow the writing of temporary files.1.168.1.100 ciscorouter(config)# ntp server 192. you’ll get errors in the /var/log/messages file: Aug 12 00:29:45 smallfry ntpd[2097]: can’t open /etc/ntp/drift. you have to set the group and owner of the directory to be ntp: [root@bigboy tmp]# chown ntp:ntp /etc/ntp If you don’t. because it is one of the most popular manufacturers of networking equipment and would feature in the overall architectures of many small office/home office (SOHO) environments and corporate departments.0 notrust nomodify notrap restrict 172.TEMP: Permission denied CONFIGURING CISCO DEVICES TO USE AN NTP SERVER You can use NTP to synchronize time on a variety of devices including networking equipment.201 ciscorouter(config)# exit ciscorouter# wr mem .CLIENT NETWORK ------#restrict 172.168. Cisco IOS To make your router synchronize with NTP servers with IP addresses 192.1. the restrict statement has only the client network defined without any keywords and the configuration line that works with other NTP versions has been commented out: # -.255. All the Fedora/Red Hat NTP daemons write temporary files to the /etc/ntp Unfortunately.1.100 and 192.qxd 2/25/05 10:16 AM Page 406 406 The NTP Server Chapter 24 In this example.16.255.16.0 Fedora Core 2 File Permissions directory.

168. CAT OS To make your router synchronize with NTP servers with IP addresses 192. 2. and Configurations. Click on the time at the bottom right of your screen. but 123 also.201. and ntp update-calendar configures the system to update its hardware clock from the software clock at periodic intervals. Unlike most UDP protocols. and set ntp client enable activates the NTP client. Click on the Internet Time tab of the dialog box 3.100 and 192. “Codes.201 ciscoswitch# exit The ntp server command forms a server association with another system. You’ll have to allow UDP traffic on source/destination port 123 between your server and the Stratum 1/2 server with which you are synchronizing. .168. Scripts. use the commands: ciscoswitch> enable password: ********* ciscoswitch# set ntp client enable ciscoswitch# ntp server 192.168. A sample Linux iptables firewall script snippet is in Appendix II. Click on the Update Now button You will get a message saying “Your time has been successfully synchronized” when the operation is complete. Click the check box labeled Automatically Synchronize with an Internet Time Server. 4.1.100 ciscoswitch# ntp server 192.1.26Harrison_ch24.qxd 2/25/05 10:16 AM Page 407 Configuring a Windows NTP Client 407 The ntp server command forms a server association with another system.168.1. and enter the name or IP address in the box underneath it. the source port isn’t a high port (above 1023).1.” CONFIGURING A WINDOWS NTP CLIENT You can add your new NTP server to your Windows client: 1. FIREWALLS AND NTP NTP servers communicate with one another using UDP with a destination port of 123.

The maintenance of firewall rules for multiple NTP connections to the Internet can also be daunting. The firewall rules can also be greatly simplified. A local NTP server is frequently a good thing to have for these reasons. A local NTP server can ensure that the clients all have the same time relative to the server even when Internet connectivity is temporarily lost. especially if the management of the firewall is handled by another group.qxd 2/25/05 10:16 AM Page 408 408 The NTP Server Chapter 24 CONCLUSION It is important that all the systems under your control have the same accurate time. thereby reducing the problems of them being out of synchronization with each other.26Harrison_ch24. Having an NTP server on your local network can make this easier to do. . It can help to give a very clear indication of a chain of events that involve multiple devices. even when they all have access there is the risk of them losing synchronization if the central connection to the Internet is lost. Sometimes it isn’t desirable for all your NTP clients to have access to the Internet to synchronize with Stratum 1 and 2 servers. and it can also help in the synchronization of time-sensitive transactions.

qxd 2/25/05 10:17 AM Page 409 P A R T III Advanced Topics 409 .27Harrison_Pt3.

qxd 2/25/05 10:17 AM Page 410 .27Harrison_Pt3.

1. Enter your selections. This chapter will briefly explain three methods of installing operating systems over a network connection. and the process can be automated. The installation procedure will then continue with the more familiar Fedora Linux installation screens. It is frequently much faster than using CDs.28Harrison_ch25.qxd 2/25/05 10:17 AM Page 411 C H A P T E R 25 Network-Based Linux Installation IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ Setting Up the Installation Server Creating a Boot CD The Network Installation Troubleshooting the Network Installation Differences Between Fedora and Red Hat Installation Automating Installation with Kickstart Conclusion Fedora Linux allows you to do operating system installations via a network connection using a Kickstart server.168. 3. 411 . HTTP. The procedure is fairly simple: 1. 2. using a single installation server (Bigboy) with an IP address of 192. Enter your preferred installation method (FTP. Connect the new server (installation client) to the same network as the server with the preloaded installation files (installation server). 4. and complete the installation. Boot the installation client from a specially created boot CD. NFS) and the needed network parameters to do this.100 as an example.

NFS. often passing through a firewall. ☞ The http:// nomenclature used by Kickstart for accessing files is more familiar to users than that used for NFS and FTP.qxd 2/25/05 10:17 AM Page 412 412 Network-Based Linux Installation Chapter 25 SETTING UP THE INSTALLATION SERVER Kickstart can be configured on an FTP. or Apache server.28Harrison_ch25. Create the Installation Directories install/Fedora/RPMS. but my experience is that the Apache server has a number of advantages over the other two. Mount your first Fedora CD-ROM: [root@bigboy tmp]# mount /dev/cdrom /mnt/cdrom 2. Unmount your CD-ROM. [root@bigboy tmp]# mkdir -p /data/network-install/Fedora/base [root@bigboy tmp]# mkdir -p /data/network-install/Fedora/RPMS [root@bigboy tmp]# mkdir -p /data/network-install/ISO You now need to place the network installation driver files into the base directory. Create the directories /data/network-install/Fedora/base. and FTP Kickstart methods all require the base set of Fedora files to be installed on the Kickstart server: 1. Copy the files from the CD-ROM base directory to the hard disk: [root@bigboy tmp]# cd /mnt/cdrom/Fedora/base [root@bigboy base]# cp -r * /data/network-install/Fedora/base 3. Each method is explained below. Strict firewall rules for HTTP are generally easier to configure than those for FTP or NFS. and use the eject command to retrieve it from the drive bay: . NFS. /data/networkand /data/network-install/ISO in which you will copy the necessary files. This may be important for you when configuring files for automated Kickstart installation. Copying the Files The HTTP. Using a Web server for Kickstart is generally easier because: ☞ Sometimes a Kickstart server has to be located on a remote network. You will place all the necessary files in the /data/network-install directory. Basic Preparation The first example sets up a Kickstart server for use in Fedora Core 2 installations.

. the code you need is: [root@bigboy tmp]# cd /data/network-install/ISO [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc1. . make sure they have the same file names as the ones you can download from the Fedora Web site. You need about 2GB of space. use: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc3. use this code to install the fourth CD: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc4.qxd 2/25/05 10:17 AM Page 413 Setting Up the Installation Server 413 [root@bigboy base]# cd /tmp [root@bigboy tmp]# umount /dev/cdrom [root@bigboy tmp]# eject cdrom You also have the option to FTP all the files from the base directory of the desired version of Fedora from the Fedora Web site to the /data/networkinstall/Fedora/base directory.iso bs=32k [root@bigboy ISO]# eject cdrom Finally.. If you create the ISO files from CDs. and place them in the /data/netdirectory. HTTP and FTP Preparation Copy all the contents of the /Fedora directory of each of the installation CDs to the /data/network-install/ directory. use: [root@bigboy ISO]# dd if=/dev/cdrom of=FC2-i386-disc2. your /data/network-install/Fedora/ directory should look like this: [root@bigboy tmp]# ls /data/network-install/Fedora/ base RPMS TRANS.iso bs=32k [root@bigboy ISO]# eject cdrom For the third. This requires about 2GB of space as well.TBL [root@bigboy tmp]# NFS Preparation work-install/ISO Create ISO images of the installation CDs.. You can download the ISO images from the Fedora Web site or use the Fedora CDs.iso bs=32k [root@bigboy ISO]# eject cdrom . When copying is complete.. [root@bigboy ISO]# eject cdrom For the second CD.iso bs=32k . For the first CD.28Harrison_ch25.

168. Here is a sample httpd.1.100/network-install/Fedora/RPMS/ or http://192.100 # # For HTTP Installations # <VirtualHost 192.1. so don’t forget the mount command.1.conf configuration: NameVirtualHost 192.168.deny allow from all </Directory> Remember to restart Apache to make these settings take effect. Here is a sample snippet of the vsftpd.168. [root@bigboy ISO]# mount /mnt/cdrom [root@bigboy ISO]# mkisofs -J -r -T -o filename. The command requires a mounted CD-ROM drive.168.iso /mnt/cdrom [root@bigboy ISO]# eject cdrom Set Up Your Web Server You will now have to set up Apache to give the file listings of your /data/network-install/Fedora and /data/network-install/ISO directories by pointing your browser to the URL http://192. respectively.100> DocumentRoot /data/ </VirtualHost> <Directory /data/network-install> Options +Indexes AllowOverride AuthConfig order allow.conf file: # # Anonymous FTP Root Directory # anon_root=/data/network-install/ # Remember to restart VSFTPD to make these settings take effect.100/network-install/Fedora/ISO/.qxd 2/25/05 10:17 AM Page 414 414 Network-Based Linux Installation Chapter 25 Note Here is a sample procedure to make ISO files with the older mkisofs command. . Set Up Your FTP Server You also have to set up your VSFTPD server to make incoming anonymous FTP connections log into the /data/network-install directory by default. You may have to install the mkisofs RPM on newer Fedora versions.28Harrison_ch25.1.

my-site. a DNS lookup on the IP address of the installation client must return a server name that maps back to the original IP address when a DNS lookup is done on that same server name.my-web-site. For example: [root@bigboy tmp]# useradd -g users ftpinstall [root@bigboy tmp]# passwd ftpinstall Changing password for user ftpinstall. nfslock. Make sure that the portmap. “Remote Disk Access with NFS.168. 3.com 192-168-1-96.168. .168. You must use tabs.192.in-addr. Create a /etc/exports file with the following entry in it.1.org has address 192.1.sync) 2. The startup scripts for these are found in the /etc/init. and netfs daemons are all running to create an NFS server. Chapter 29. [root@bigboy tmp]# [root@bigbot tmp]# usermod -d / ftpinstall [root@bigbot tmp]# Set Up Your NFS Server The steps for setting up an NFS server are more complicated: 1. You must also make sure that the user has read access to the /data/network-install directory. New password: Retype new password: passwd: all authentication tokens updated successfully. For example: [root@bigboy tmp]# host 192.28Harrison_ch25.” explains this in more detail.96 [root@bigboy tmp]# This may mean that you must create entries for all your DHCP IP addresses if you choose to use a DHCP method of assigning IP addresses during installation.1. [root@bigboy tmp]# [root@bigboy tmp]# host 192-168-1-96.96 96.myweb-site. not spaces between the entries: /data/network-install *(ro. nfs.qxd 2/25/05 10:17 AM Page 415 Setting Up the Installation Server 415 Create a Special FTP User You can create a special user for non-anonymous FTP installations with the user’s home directory as /.org. In other words.d directory. Run the exportfs command to add this directory to the NFS database of network available directories: [root@bigboy tmp]# exportfs -ra Keep in mind that the installation client must have a matching pair of forward and reverse DNS entries on your DNS server.arpa domain name pointer 192-168-1-96.

168. It also reduces the possibility of human error. Boot your system using the first Fedora installation CD. Enter the command linux askmethod at the boot: prompt 4. From the list of choices Local CDROM Hard Drive NFS Image FTP HTTP select the network option of your choice (NFS. FTP. Enter the IP address of the installation server for NFS Server Name. If you selected the NFS method. The remaining menus will be the usual Fedora GUI installation screens. Setting up the Installation Server as a DHCP server is fairly straight forward. 7. you now come to the NFS Setup menu. the installation procedure mimics the regular Linux installation. Go through the usual steps until prompted for the installation method. Press the Enter key. “Configuring the DHCP Server. THE NETWORK INSTALLATION From here on. 8. 2. 3.28Harrison_ch25. This would most likely be interface eth0.” CREATING A BOOT CD You will need the first CD in your set of Fedora CDs prior to actually starting the installation.1.100. or HTTP).iso file onto a CD. the installation client prompts you for the IP address it should use for the installation process. Select the Ethernet device with which the installation client is connected to the installation server network. For Red Hat Directory specify /data/network-install/ISO. 6.qxd 2/25/05 10:17 AM Page 416 416 Network-Based Linux Installation Chapter 25 Configure Your DHCP Server During the installation procedure. Connect your client Linux box to the DHCP network. This automates the installation more. . except for the first couple steps: 1. for details see Chapter 8. 5. You can either copy an existing disk or burn a copy of the FC2-i386-disc1. Select DHCP in the Configure TCP/IP screen. thereby making it faster. making the Installation client use DHCP during the installation. This is the only CD you’ll need for future network installations. I recommend selecting the option that makes the installation client get its address via DHCP. in this case 192.

which can be useful if you have a large number of servers to install.28Harrison_ch25. TROUBLESHOOTING THE NETWORK INSTALLATION You can do some basic troubleshooting by accessing the various installation status screens available: ☞ ☞ ☞ ☞ To view the installation logs. and the remaining menus will be text-based versions of the usual Fedora installation screens. If you do select the NonAnonymous FTP box. press Ctrl+Alt+F4. you are given the FTP Setup menu. The remaining menus will be textbased versions of the usual Fedora installation screens. Enter the username and password of your special FTP user account. press Ctrl+Alt+F1 from text-based installations or Ctrl+Alt+F7 from the GUI.168.168.100 again). but Red Hat expects the files to be placed in a directory named RedHat under /data/network-install. If you selected the FTP method. This is the only difference between both operating systems when configuring network installations.cfg Kickstart configuration file. To access a limited BASH shell kernel. Enter the IP address of the installation server as the FTP site name (192. Fedora expects the files to be placed in a directory named Fedora under /data/network-install.1.100. If you do not select the Non-Anonymous FTP box. . Do not add an http:// to the beginning. then the Red Hat directory is /data/network-install.1. you are shown the HTTP Setup menu. To see kernel messages. DIFFERENCES BETWEEN FEDORA AND RED HAT INSTALLATION The RPM file locations have to be different when using the HTTP and FTP methods. press Ctrl+Alt+F2. Enter the IP address of the installation server when prompted for a Web site name. You can use this file to create an automated installation of a duplicate system. press Ctrl+Alt+F3 at any time. in this case 192. To return to the main installation screen at any time. the Red Hat directory is simply /.qxd 2/25/05 10:17 AM Page 417 Automating Installation with Kickstart 417 If you selected the HTTP method. The Red Hat directory is /network-install. The remaining menus will be text-based versions of the usual Fedora installation screens. AUTOMATING INSTALLATION WITH KICKSTART Both Fedora and Red Hat Linux save all the parameters you used during installation in the /root/anaconda-ks.

How to Create New Kickstart Configuration Files You can create a customized Kickstart configuration file by using the ksconfig command from a GUI console. Note The IP address you assign must be on the same subnet as that of the DHCP server for Kickstart to work. For example: %post chkconfig chkconfig chkconfig chkconfig chkconfig chkconfig isdn off pcmcia off portmap off apmd off nfslock off nfs off A Note about Using anaconda-ks. You can use HTTP and FTP. you save the configuration with the filename of your choice. Adding post Installation Commands You may want to run some commands on the newly created Linux installation after Kickstart is complete.28Harrison_ch25. If the server is going to reside on a different network after the installation.qxd 2/25/05 10:17 AM Page 418 418 Network-Based Linux Installation Chapter 25 This section shows you how to automate network installations using the Kickstart application and NFS. then you’ll have to run a separate script to change the IP addressing information after the installation is complete. It brings up a menu from which you can select all your installation options. Fedora comments out the partitioning information in this file. You may want to then edit the configuration file. Note Do not change the order of the entries in the Kickstart configuration file. You can disable them by placing a %post section at the end of the Kickstart file with all the post installation commands you wish to run. When finished.cfg file as a template for future installations. Some processes that Fedora activates by default may not be suitable for your server and may need to be disabled. During the Kickstart process you will be prompted for these unspecified values. but I omitted them to keep the discussion brief. including the system’s name and IP address.cfg It is possible to use the /root/anaconda-ks. adding # signs to comment out certain parameters that may change from system to system. so .

1. If not. How to Run a Kickstart Installation work-install directory. For the NFS method.100:/kickstart/ks. Manually Specifying the Kickstart Filename After you boot from your boot CD-ROM. The ks.100 --dir=/data/network-install/ISO If they don’t you may be prompted for NFS ISO file location information.qxd 2/25/05 10:17 AM Page 419 Automating Installation with Kickstart 419 you have to uncomment it and then make your partitioning modifications or be prepared to be prompted for your portioning information. The upcoming examples assume called /data/network-install/kickstart.16. you need to issue a command at the boot: prompt to continue with the installation.cfg file is the Kickstart configuration file you want to use.168. Booting With Your Kickstart Files There are two ways to specify the name of the Kickstart file to use.28Harrison_ch25. but the command syntax changes depending on whether you choose NFS or HTTP. use: boot: linux ks=http://192. The second is to have your DHCP server automatically tell the Kickstart client the name of the Kickstart file when the server assigns the IP address.1. When using a Web server.1. verify that the first two lines of the file look like this: install url --url http://192. Remember that It is best to place your Kickstart files in a subdirectory under the /data/netthe subdirectory is you may want to remove the # comments from the partition section of the file. The first is to enter it manually from the boot: prompt when you insert the boot CD. When using a NFS server. verify that the first two lines of the file look like this: install nfs --server=192.1.168.100/network-install/ If they don’t.cfg For the HTTP method.100/network-install/kickstart/ks. you will be prompted for this information. you may be prompted for RPM base file location information.cfg . use: boot: linux ks=nfs:192.168.

Place your Kickstart file in the /data/network-install/kickstart directory. next-server 192. 2.1. Insert the boot CD into the Kickstart client Linux box.168.28Harrison_ch25. 3. but you may want to research this option some more.qxd 2/25/05 10:17 AM Page 420 420 Network-Based Linux Installation Chapter 25 Configuring the Filename Automatically Whenever you have to create lots of cloned Linux servers.cfg”. Edit your dhcpd.conf file. It then automatically attempts to get a DHCP IP address and see if the DHCP server will specify a configuration file. Here is how it’s done: 1. A Kickstart server connected to an isolated wireless network dedicated to the purpose may be a good idea for data centers with hundreds of Linux servers. if you have more complex projects. filename “/data/network-install/kickstart/ks. The next-server value is the IP address of the Kickstart server. The rest should be automatic. Kickstart next uses NFS to get both the configuration file and the installation ISOs. but also because it can be left unattended to install a predetermined Linux configuration. you may want to configure your DHCP server to specify the single Kickstart configuration file you wish to use. .100. The topic is beyond the scope of this book. Time is saved not only because a network connection can be faster than using CDs. A recent standard called PXE enables you to run Kickstart without a CD-ROM if you configure the NIC to do a network boot from a specially configured DHCP server. CONCLUSION The Kickstart method of Fedora Linux installation can greatly reduce the length of time it takes to install the operating system. and add the following lines to the section for the interface that will be serving DHCP IP addresses.cfg on the boot CD. At the boot: prompt type: boot: linux ks Kickstart first searches for a configuration file named ks. and connect it to the DHCP network.

Data is written to the next device in the chain only after the previous one is filled. The aim of Linear RAID is to accommodate large filesystems spread over multiple devices with no data redundancy. RAID can be configured using a variety of standards.qxd 2/25/05 10:18 AM Page 421 C H A P T E R 26 Linux Software RAID IN THIS CHAPTER ☞ ☞ ☞ ☞ RAID Types Before You Start Configuring Software RAID Conclusion The main goals of using redundant arrays of inexpensive disks (RAID) are to improve disk data performance and provide data redundancy.or software-based. RAID can be handled either by the operating system software or it may be implemented via a purpose built RAID disk controller card without having to configure the operating system at all.29Harrison_ch26. 421 . Linear Mode RAID In the Linear RAID. Linear mode RAID is not supported by Fedora Linux. the RAID controller views the RAID set as a chain of disks. For the sake of simplicity. Take a look at the most popular. This chapter will explain how to configure the software RAID schemes supported by Red Hat/Fedora Linux. RAID TYPES Whether hardware. the chapter focuses on using RAID for partitions that include neither the /boot or the root (/) filesystems. A drive failure will corrupt your data.

cherry and strawberry—and four plates. vanilla.1 illustrates the data allocation process in RAID 0. . Figure 26. and think of the data as a cake. RAID 0 is supported by Fedora Linux. it then continues the striping using the available space on the remaining drives. Like Linear RAID. four 9GB hard disks configured in a RAID 0 set are seen by the operating system to be one 36GB disk. because the total number of RAID drives available is reduced. RAID 0 aims to accommodate large filesystems spread over multiple devices with no data redundancy. When this occurs. When RAID runs out of striping space on the smallest device.29Harrison_ch26. For example.qxd 2/25/05 10:18 AM Page 422 422 Linux Software RAID Chapter 26 RAID 0 With RAID 0. the RAID controller tries to evenly distribute data across all disks in the RAID set. A file that is spread over four disks can be read four times as fast.1 RAID 0 and RAID 1 operation. Envision a disk as if it were a plate. RAID 0 is best used with drives of equal size. Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8 Logical Drive Disk Striping (RAID 0) Block 1 Block 3 Block 5 Block 7 Physical Disk #1 Block 1 Block 2 Block 3 Block 4 Block 2 Block 4 Block 6 Block 8 Physical Disk #2 Block 5 Block 6 Block 7 Block 8 Physical Disk #1 Disk Mirroring (RAID 1) Block 1 Block 2 Block 3 Block 4 Block 5 Block 6 Block 7 Block 8 Physical Disk #2 Figure 26. The RAID 0 drivers make the operating system feel that the cakes are intact and placed on one large plate. the data access speed is lower for this portion of data. You have four cakes—chocolate. The initialization process of RAID 0 divides the cakes and distributes the slices across all the plates. For this reason. The advantage of RAID 0 is data access speed. RAID 0 can accommodate disks of unequal sizes. You should also be aware that RAID 0 is often called striping.

RAID 4 isn’t used very frequently. A disadvantage of software-based RAID 1 is that the server has to send data twice to be written to each of the mirror disks. any data written to any section of the data portion of the RAID set must be followed by an update of the parity disk.29Harrison_ch26. it is repopulated with the lost data with the help of the parity disk’s information. When the failed disk is replaced. RAID 4 is not supported by Fedora Linux. RAID 4 RAID 4 operates like RAID 0 but inserts a special error-correcting or parity chunk on an additional disk dedicated to this purpose. For this reason. Unlike RAID 0. without the speed advantages of RAID 0. This avoids the parity disk bottleneck. while maintaining many of the speed features of RAID 0 and the redundancy of RAID 1. The parity disk can therefore act as a bottleneck. the other one continues to function. RAID 1 also offers the possibility of using a hot standby spare disk that will be automatically cloned in the event of a disk failure on any of the primary RAID devices. but the parity information is not. RAID 4 combines the high speed provided of RAID 0 with the redundancy of RAID 1. the data is automatically cloned to the new disk from the surviving disk. data is cloned on a duplicate disk. When the failed disk is replaced. When this occurs. When one of the disks in the RAID set fails. This can saturate data busses and CPU use. RAID 1 is supported by Fedora Linux. Figure 26. RAID 5 can survive the loss of a single disk only. Think of telling two people the same story so that if one forgets some of the details you can ask the other one to remind you. and the disk controller then duplicates the data to the mirror disks. the data in it can be recreated on the fly with the aid of the information on the RAID set’s parity disk. RAID 1 offers data redundancy. .qxd 2/25/05 10:18 AM Page 423 RAID Types 423 RAID 1 With RAID 1.1 illustrates the data allocation process in RAID 1. Like RAID 4. This makes RAID-capable disk controllers the preferred solution when implementing RAID 1. RAID 5 RAID 5 improves on RAID 4 by striping the parity data between all the disks in the RAID set. the server CPU sends the data to the RAID disk controller once. With a hardware-based solution. the extra space on the larger device isn’t used. Its major disadvantage is that the data is striped. A limitation of RAID 1 is that the total RAID size in gigabytes is equal to that of the smallest disk in the RAID set. RAID 4 requires at least three disks in the RAID set and can survive the loss of a single drive only. This RAID method is therefore frequently called disk mirroring. In other words.

Figure 26. then you should be aware of these basic guidelines to follow when setting up software RAID. Linux RAID 5 requires a minimum of three disks or partitions.29Harrison_ch26.2 RAID 5 operation. Block 1 Block 2 Block 3 Block 1 Block 4 Block 5 Block 6 Block 6 Block 7 Block 8 Logical Drive Block 7 Block 8 Error Check Blocks 7 and 8 Physical Disk #3 Error Check Blocks 3 and 4 Block 3 Block 4 Block 2 Error Check Blocks 1 and 2 Error Check Blocks 5 and 6 Block 5 Physical Disk #1 Physical Disk #2 RAID 5 Figure 26. BEFORE YOU START Specially built hardware-based RAID disk controllers are available for both IDE and SCSI drives. IDE Drives To save costs. many small business systems will probably use IDE disks. They usually have their own BIOS. Hardware-based RAID is transparent to your operating system.2 illustrates the data allocation process in RAID 5. If hardware RAID isn’t available.qxd 2/25/05 10:18 AM Page 424 424 Linux Software RAID Chapter 26 RAID 5 is supported by Fedora Linux. but they do have some limitations: . the hardware does all the work. so you can configure them right after your system’s power on self test (POST).

Serial ATA Drives Serial ATA type drives are rapidly replacing IDE. which makes hot swapping or the ability to replace components while the system is still running. For these reasons. especially in a corporate environment. possible without the fear of affecting other devices on the data cable. I recommend you use only one IDE drive per controller when using RAID. .qxd 2/25/05 10:18 AM Page 425 Before You Start 425 ☞ The total length of an IDE cable can be only a few feet long. IDE-based software RAID may be adequate. or ultra ATA. ☞ The performance of the IDE bus can be degraded by the presence of a second device on the cable. ☞ The failure of one drive on an IDE bus often causes the malfunctioning of the second device. which generally limits IDE drives to small home systems. ☞ Serial ATA has better error checking than IDE. ☞ There is only one drive per cable. ☞ Only two devices can be attached per controller. SCSI Drives SCSI hard disks have a number of features that make them more attractive for RAID use than either IDE or Serial ATA drives: ☞ SCSI controllers are more tolerant of disk failures. If you can’t afford more expensive and faster SCSI drives. which makes them easier to configure. You cannot replace them while your system is running. The failure of a single drive is less likely to disrupt the remaining drives on the bus. ☞ IDE drives do not hot swap. ☞ IDE drives have a 133Mbytes/s data rate whereas the Serial ATA specification starts at 150 Mbytes/sec with a goal of reaching 600 Mbytes/s over the expected ten year life of the specification. drives as the preferred entry-level disk storage option because of a number of advantages: ☞ The drive data cable can be as long as one meter in length versus IDE’s 18 inches. Serial ATA would be the preferred device for software and hardware RAID. This can be fatal if you have two IDE drives of the same RAID set attached to the same cable.29Harrison_ch26. ☞ There are no jumpers to set on Serial ATA drives to make it a master or slave. In a home or SOHO setting.

☞ SCSI currently supports data rates of up to 640 Mbytes/s making them highly desirable for installations where rapid data access is imperative. It can accommodate 7 (single-ended SCSI) or 15 (all other SCSI types) devices. issue the exit command. all the partitions on each RAID disk should be part of a RAID set. ☞ Much more than two devices may be connected to a SCSI cable bus. and your system will boot in the default runlevel provided in the /etc/inittab file. Back up Your System First Software RAID creates the equivalent of a single RAID virtual disk drive made up of all the underlying regular partitions used to create it. you should also consider configuring RAID while your system is running in single-user mode from the VGA console. ☞ Some models of SCSI devices support hot swapping.29Harrison_ch26. which allows you to replace them while the system is running. SCSI drives tend to be more expensive than IDE drives. Configure RAID in Single-User Mode As you will be modifying the disk structure of your system. making them suitable for data center applications. The reason for this is obvious: A disk failure could still incapacitate a system. A mistake could unintentionally corrupt valid data. Formatting. however.qxd 2/25/05 10:18 AM Page 426 426 Linux Software RAID Chapter 26 ☞ SCSI cables can be up to 25 meters long. which may make them less attractive for home use. This makes sure that most applications and networking are shutdown and that no other users can access the system. Many people simplify this problem by filling each disk of a RAID set with only one partition. . reducing the risk of data corruption during the exercise. It is best to back up the data on these and any other partitions on the disk drive on which you want implement RAID. If you decide to use RAID. however. [root@bigboy tmp]# init 1 Once finished. causes all the old data on the underlying RAID partitions to be lost. You have to format this new RAID device before your Linux system can store files on it. Should I Use Software RAID Partitions or Entire Disks? It is generally not a good idea to share RAID-configured partitions with nonRAID partitions.

63 sectors/track. each on a separate disk.29Harrison_ch26.qxd 2/25/05 10:18 AM Page 427 Configuring Software RAID 427 CONFIGURING SOFTWARE RAID Configuring RAID using Fedora Linux requires a number of steps that need to be followed carefully. which shows only mounted filesystems but has the big advantage of giving you the mount points too. You may then also want to use the df -k command.. /dev/hda15 1455 [root@bigboy tmp]# End 13 144 209 1467 655 Blocks 104391 1052257+ 522112+ 10104885 3582463+ Id 83 83 82 5 83 System Linux Linux Linux swap Extended Linux 1467 104391 83 Linux . RAID limits the extent of data access on each partition to an area no larger than that of the smallest partition in the RAID set. In the tutorial example. 1467 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start /dev/hda1 * 1 /dev/hda2 14 /dev/hda3 145 /dev/hda4 210 /dev/hda5 210 . These two commands should help you to easily identify the partitions you want to use. Determining Available Partitions First use the fdisk -l command to view all the mounted and unmounted filesystems available on your system.. as in this scenario.0 GB. 12072517632 bytes 255 heads. . RAID Partitioning You first need to identify two or more partitions.. Here is some sample output of these commands: [root@bigboy tmp]# fdisk –l Disk /dev/hda: 12.. the partitions should be of approximately the same size. you’ll be configuring RAID 5 using a system with three pre-partitioned hard disks. The partitions to be used are: ☞ ☞ ☞ /dev/hde1 /dev/hdf2 /dev/hdg1 Be sure to adapt the various stages outlined below to your particular environment. If you are doing RAID 0 or RAID 5.

DOS FDISK. OS/2 FDISK) Command (m for help): Use fdisk Help Now use the fdisk m command to get some help: Command (m for help): m . There is nothing wrong with that. ... . so you need to make sure they are unmounted: [root@bigboy tmp]# umount /dev/hde1 [root@bigboy tmp]# umount /dev/hdf2 [root@bigboy tmp]# umount /dev/hdg1 Prepare the Partitions with fdisk You have to change each partition in the RAID set to be of type FD (Linux raid autodetect)..... and you can do this with fdisk.. old versions of LILO) 2) booting and partitioning software from other OSs (e. print the partition table quit without saving changes create a new empty Sun disklabel change a partition’s system id Command (m for help): .. and could in certain setups cause problems with: 1) software that runs at boot time (e.. /dev/hda7 [root@bigboy tmp]# 1K-blocks 1035692 101086 101086 163916 8357 4127 Used Available Use% 819164 87510 91740 17% / 9% /boot 5% /data1 5336664 464228 4601344 10% /var Unmount the Partitions You don’t want anyone else accessing these partitions while you are creating the RAID set.g..qxd 2/25/05 10:18 AM Page 428 428 Linux Software RAID Chapter 26 [root@bigboy tmp]# df –k Filesystem Mounted on /dev/hda2 /dev/hda1 /dev/hda15 .. p q s t . Here is an example using /dev/hde1: [root@bigboy tmp]# fdisk /dev/hde The number of cylinders for this disk is set to 8355.g.. .29Harrison_ch26... ... but this is larger than 1024.

. 16 Hidden FAT16 17 Hidden HPFS/NTF 18 AST SmartSleep 1b Hidden Win95 FA Hex code (type L to Changed system type 61 SpeedStor a9 NetBSD f2 DOS secondary 63 GNU HURD or Sys ab Darwin boot fd Linux raid auto 64 Novell Netware b7 BSDI fs fe LANstep 65 Novell Netware b8 BSDI swap ff BBT list codes): fd of partition 1 to fd (Linux raid autodetect) Command (m for help): Make Sure the Change Occurred Use the p command to get the new proposed partition table: Command (m for help): p Disk /dev/hde: 4311 MB.qxd 2/25/05 10:18 AM Page 429 Configuring Software RAID 429 Set the ID Type to FD Partition /dev/hde1 is the first partition on disk /dev/hde. You also should use the L command to get a full listing of ID types in case you forget: Command (m for help): t Partition number (1-5): 1 Hex code (type L to list codes): L . Modify its type using the t command.... 4311982080 bytes 16 heads..29Harrison_ch26. .. .. and specify the partition number and type code. 63 sectors/track. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde4 /dev/hde5 /dev/hde6 Start 1 4089 6608 6608 7501 End 4088 5713 8355 7500 8355 Blocks 2060320+ 819000 880992 450040+ 430888+ Id fd 83 5 83 83 System Linux raid autodetect Linux Extended Linux Linux Command (m for help): Save the Changes Use the w command to permanently save the changes to disk /dev/hde: Command (m for help): w The partition table has been altered! Calling ioctl() to re-read partition table.

1. To ensure success. Here. and 5. remember these general guidelines: ☞ When configuring RAID 5. if you have four partitions for RAID. The data is distributed across the drives in 32MB chunks: # # sample raiddev configuration file # ‘old’ RAID0 array created with mdtools. Edit the RAID Configuration File The Linux RAID configuration file is /etc/raidtab. 2. ☞ The raid-disk parameters for each partition in the /etc/raidtab file are numbered starting at 0. you must use a parity-algorithm setting. For an explanation of the various parameters. RAID 5 is configured to use each of the desired partitions on the three disks. [root@bigboy tmp]# The error above will occur if any of the other partitions on the disk is mounted. It’s enough to know that the steps for changing the IDs for /dev/hdf2 and /dev/hdg1 are very similar. and the set of three is called /dev/md0. The kernel still uses the old table. the /etc/raidtab persistent-superblock must be set to 1 for the RAID autodetect feature (partition type FD) to work. You can find templates for this file in the /usr/share/doc/raidtools* directory. Consider an example. For all other RAID versions. # raiddev /dev/md0 raid-level 5 nr-raid-disks 3 persistent-superblock 1 chunk-size 32 parity-algorithm left-symmetric device /dev/hde1 raid-disk 0 device /dev/hdf2 raid-disk 1 device /dev/hdg1 raid-disk 2 . ☞ For RAID levels 1. issue the man raidtab command.qxd 2/25/05 10:18 AM Page 430 430 Linux Software RAID Chapter 26 WARNING: Re-reading the partition table failed with error 16: Device or resource busy. Syncing disks. they would be numbered 0.29Harrison_ch26. Repeat for the Other Partitions For the sake of brevity. 4. The new table will be used at the next reboot. For example. persistent-superblock must be set to 0. and 3. I won’t show the process for the other partitions.

32768 fragments per group 16128 inodes per group . algorithm 3 [3/3] [UUU] unused devices: <none> [root@bigboy tmp]# Format the New RAID Set Your new RAID device now has to be formatted. raid superblock at 104320kB [root@bigboy tmp]# Confirm RAID Is Correctly Initialized The /proc/mdstat file provides the current status of all RAID devices. 1030160 blocks 51508 blocks (5. [root@bigboy tmp]# mkraid /dev/md0 analyzing super-block disk 0: /dev/hde1. which is comprised of 8 blocks. 104391kB. If the values don’t match.00%) reserved for the super user First data block=0 32 block groups 32768 blocks per group. 104391kB. Confirm that the initialization is finished by inspecting the file and making sure that there are no initialization-related messages: [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors md0 : active raid5 hdg1[2] hde1[1] hdf2[0] 4120448 blocks level 5.32 (09-Nov-2002) Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) 516096 inodes. Here a block size of 4KB (4096 bytes) is used with each chunk. raid superblock at 104320kB disk 2: /dev/hdg1. you will get parity errors. raid superblock at 104320kB disk 1: /dev/hdf2. 104391kB. It is very important that the chunk-size parameter in the /etc/raidtab file match the value of the block size multiplied by the stride value in the command below. [root@bigboy tmp]# mke2fs -j -b 4096 -R stride=8 /dev/md0 mke2fs 1.29Harrison_ch26.qxd 2/25/05 10:18 AM Page 431 Configuring Software RAID 431 Create the RAID Set The mkraid command creates the RAID set by reading the The example creates the logical RAID device /dev/md0: /etc/raidtab file. 32k chunk. The next example uses the -j qualifier to ensure that a journaling filesystem is created.

294912. [root@bigboy tmp]# Load the RAID Driver for the New RAID Set Next. 229376. Device mounting then occurs later on in the boot process.d/rc.29Harrison_ch26. 884736 Writing inode tables: done Creating journal (8192 blocks): done Writing superblocks and filesystem accounting information: done This filesystem will be automatically checked every 26 mounts or 180 days. Mounting a RAID device that doesn’t have a loaded driver can corrupt your data and produce this error: Starting up RAID devices: md0(skipped) Checking filesystems .sysinit script checks the /etc/fstab file for device entries that match RAID set names in the /etc/raidtab file. The script will not automatically start the RAID set driver for the RAID set if it doesn’t find a match. whichever comes first. 819200. On startup. 98304. Add an Entry for the RAID set. 163840. create a mount point for called /mnt/raid: [root@bigboy mnt]# mkdir /mnt/raid /dev/md0. the /etc/rc. just use the real device name. make the Linux operating system fully aware of the RAID set by loading the driver for the new RAID set using the raidstart command: [root@bigboy tmp]# raidstart /dev/md0 [root@bigboy tmp]# Create a Mount Point for the RAID Set After the driver loads. the /dev/md0 device: /dev/md0 /mnt/raid ext3 defaults 1 2 Do not use labels in the /etc/fstab file for RAID devices.qxd 2/25/05 10:18 AM Page 432 432 Linux Software RAID Chapter 26 Superblock backups stored on blocks: 32768. such as this one Edit the /etc/fstab File The /etc/fstab file lists all the partitions that need to mount when the system boots. such as /dev/md0. Use tune2fs -c or -i to override.

This command is run automatically at boot time. [root@bigboy tmp]# mount /dev/md0 /mnt/raid . You have your choice of methods: ☞ The mount command’s -a flag causes Linux to mount all the devices in the /etc/fstab file that have automounting enabled (default) and that are also not already mounted. Run fsck manually (ie without -a or -p options).qxd 2/25/05 10:18 AM Page 433 Configuring Software RAID 433 /raiddata: Superblock has a bad ext3 journal(inode8) CLEARED. so you’ll only have to do this once. The physical size of the device is 8960245 blocks. [root@bigboy tmp]# raidstart /dev/md0 Mount the New RAID Set Use the mount command to mount the RAID set. [root@bigboy tmp]# mount -a ☞ You can also mount the device manually. The /dev/hde1. 41/26104 files. 12755/104391 blocks /raiddata: UNEXPECTED INCONSISTENCY. Make sure that all references to them in this file are commented with a # at the beginning of the line or deleted entirely: #/dev/hde1 #/dev/hdf2 #/dev/hdg1 /data1 /data2 /data3 ext3 ext3 ext3 defaults defaults defaults 1 2 1 2 1 2 Start the New RAID Set’s Driver You now can start the new RAID set’s driver with the raidstart command. If you are not familiar with the /etc/fstab file. and /dev/hdg1 partitions were replaced by the combined /dev/md0 partition. You therefore don’t want the old partitions to be mounted again. ***journal has been deleted .file system is now ext 2 only*** /raiddata: The filesystem size (according to the superblock) is 2688072 blocks. use the man fstab command to get a comprehensive explanation of each data column it contains. /dev/hdf2. Either the superblock or the partition table is likely to be corrupt! /boot: clean.29Harrison_ch26.

[root@bigboy tmp]# raidstart /dev/md0 [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors md0 : active raid5 hdg1[2] hde1[1] hdf2[0] 4120448 blocks level 5. as seen here: [root@bigboy tmp]# raidstop /dev/md0 [root@bigboy tmp]# cat /proc/mdstat Personalities : [raid5] read_ahead 1024 sectors unused devices: <none> [root@bigboy tmp]# More information. Unlike software RAID. Remember to take these factors into consideration when determining the right solution for your needs and research the topic thoroughly before proceeding. it tends to be expensive.29Harrison_ch26.qxd 2/25/05 10:18 AM Page 434 434 Linux Software RAID Chapter 26 Check the Status of the New RAID The /proc/mdstat file provides the current status of all the devices. the file has very little information. hardware RAID requires entire disks to be dedicated to the purpose and when combined with the fact that it usually requires faster SCSI hard disks and an additional controller card. 32k chunk. but it tends to be slower and less reliable than RAID provided by a hardwarebased RAID disk controller. When the RAID driver is stopped. . algorithm 3 [3/3] [UUU] unused devices: <none> [root@bigboy tmp]# CONCLUSION Linux software RAID provides redundancy across partitions and hard disks. including the partitions of the RAID set. Hardware RAID configuration is usually done via the system BIOS when the server boots up. Weighing cost versus reliability is always a difficult choice in systems administration. and once configured. is provided after you load the driver using the raidstart command. it is absolutely transparent to Linux.

Scenario Things are getting crowded on bigboy: Even after you removed all unwanted data. The first is by moving directories from a full partition to an empty one made available by the new disk and then linking the directory structures of the two disks together. To provide more space. ADDING DISKS TO LINUX At some stage you’ll be faced with the task of installing an additional hard drive into your Linux server. The most common reasons for this are expanding databases. this section will cover adding a hard disk with only one partition and will then explain how to migrate data from the full disk to the new one. The second is by merging the partitions together to create a combined partition using the Linux Logical Volume Manager (LVM). Perhaps an existing device failed. increasing numbers of users. and the larger number of tasks your Linux server is expected to perform until a replacement is found.qxd 2/25/05 10:18 AM Page 435 C H A P T E R 27 Expanding Disk Capacity IN THIS CHAPTER ☞ Adding Disks to Linux ☞ Expanding Partitions with LVM ☞ Conclusion The lack of available disk storage frequently plagues Linux systems administrators. the /var partition would still be over 95% full. This chapter explores how to add a disk to a Linux system in two ways.30Harrison_ch27. or maybe you ran out of available space. You need to add a new 435 .

the third would be c. and so on. the system has two hard disks. Finally. You can verify this situation with the df -k command’s output.30Harrison_ch27. SCSI disks all start with sd and IDE disks with hd.qxd 2/25/05 10:18 AM Page 436 436 Expanding Disk Capacity Chapter 27 hard drive to the system. [root@bigboy tmp]# cat /proc/partitions major minor #blocks name 3 0 7334145 3 1 104391 3 2 1052257 3 3 2040255 3 4 1 3 5 3582463 3 6 554211 22 0 78150744 [root@bigboy tmp]# hda hda1 hda2 hda3 hda4 hda5 hda6 hdb Note Linux hard disk device names follow a specific standard. . a two-digit number defines the partition number. The entire hard disk is represented by an entry with a minor number of 0. Using this convention the fifth partition on the fourth IDE drive would be /dev/hdd5. but you now need to know how to proceed. In the example. After this comes a letter that identifies the unit number of the disk. but the new disk (/dev/hdb) needs to be prepared to accept data. which also shows that the other partitions are too full to accept any more data: [root@bigboy tmp]# df -k Filesystem 1K-blocks /dev/hda3 505636 /dev/hda1 101089 none 63028 /dev/hda5 248895 /dev/hda7 3304768 /dev/hda2 3304768 [root@bigboy tmp]# Used Available Use% Mounted on 118224 361307 25% / 14281 81589 15% /boot 0 63028 0% /dev/shm 6613 229432 3% /tmp 2720332 416560 87% /usr 3300536 4232 99% /var A new hard disk was added according to the manufacturer’s instructions. and all the partitions on the drive are sequentially numbered after that. the first disk would be a. so for example. disk /dev/hda has been partitioned. Determining the Disk Types Linux stores the names of all known disk partitions in the /proc/partitions file. the second would be b.

Type the fdisk command followed by the name of the disk. The first Linux step in adding a new disk is to partition it in preparation of adding a filesystem to it. default 9729): 4. OS/2 FDISK) Command (m for help): 2. The steps are: 1. 80026361856 bytes 255 heads. and could in certain setups cause problems with: 1) software that runs at boot time (e. there are none. which is good: Command (m for help): p Disk /dev/hdb: 80.. You will see that n is the command to add a new partition.. because both operating systems share the fdisk partitioning utility. issue the p command to print all the known partitions on the disk. Run the print (p) command to confirm that you successfully created the partition.g. Command (m for help): p Disk /dev/hdb: 80.g.0 GB. but this is larger than 1024.0 GB. default 1):<RETURN> Using default value 1 Last cylinder or +size or +sizeM or +sizeK (1-9729.30Harrison_ch27. 63 sectors/track. There is nothing wrong with that. so the command is [root@bigboy tmp]# fdisk /dev/hdb The number of cylinders for this disk is set to 9729. In this case. The fdisk m command prints a small help manual of valid commands.qxd 2/25/05 10:18 AM Page 437 Adding Disks to Linux 437 Preparing Partitions on New Disks Linux partition preparation is very similar to that in a Windows environment. number 1. You want to run fdisk on the /dev/hdb disk. DOS FDISK. Just to make sure you’re on the correct device. Add a new primary partition. and use the defaults to make the partition occupy the entire disk: Command (m for help): n Command action e extended p primary partition (1-4) p Partition number (1-4): 1 First cylinder (1-9729. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot Start End Blocks Id System Command (m for help): 3. old versions of LILO) 2) booting and partitioning software from other OSs (e. 80026361856 bytes .

Command (m for help): w Command (m for help): q . you can convert one of the primary ones into an extended one. such as 82 for swap space. The t command enables you to change partition type from the default of 83 for regular Linux partitions to something else. Here is an example of a partition table that includes an extended partition followed by two regular partitions within it: Command (m for help): p Disk /dev/hda: 7510 MB. when finished. In most cases. the default value is sufficient. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hdb1 Start 1 End 9726 Blocks 78148161 Id 83 System Linux Command (m for help): Tip If you make a mistake. 913 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hda1 * /dev/hda2 /dev/hda3 /dev/hda4 /dev/hda5 /dev/hda6 Start 1 14 145 399 399 845 End 13 144 398 913 844 913 Blocks 104391 1052257+ 2040255 4136737+ 3582463+ 554211 Id 83 83 82 5 83 83 System Linux Linux Linux swap Extended Linux Linux Command (m for help): Adding more partitions is just a question of repeating the previous steps the required number of times. Changes won’t be made to the disk’s partition table until you use the w command to write. you may have noticed that fdisk queried you as to whether it was going to be a primary or secondary partition. 5. while remembering that at some stage. this won’t be necessary. 63 sectors/track. Note When you created the new partition. 7510164480 bytes 255 heads. you may need to add an extended partition. you can use the d command to delete the partition and start over. 63 sectors/track. exit with the q command. the changes.30Harrison_ch27. Do that now. Linux allows only four primary partitions. if you need more. or save. and.qxd 2/25/05 10:18 AM Page 438 438 Expanding Disk Capacity Chapter 27 255 heads.

30Harrison_ch27.. . giving it a new directory structure by using the mkfs command.. Create directory /mnt/hdb1 for this purpose: [root@bigboy tmp]# mkdir /mnt/hdb1 . you’ll need to verify your work and start migrating your data to the new disk.. These steps will be covered next.0 GB. which is what you should use here: [root@bigboy tmp]# mkfs -t ext3 /dev/hdb1 Next. . The Fedora installation procedure defaults to an ext3 type... you must create a special mount point directory.... .qxd 2/25/05 10:18 AM Page 439 Adding Disks to Linux 439 After this is complete.. Verifying Your New Partition You can take a look at the /proc/partitions file or use the fdisk to see the changes to the disk partition structure of your system: [root@bigboy major minor . 80026361856 bytes 255 heads. Disk /dev/hdb: 80.. 9729 cylinders Units = cylinders of 16065 * 512 = 8225280 bytes Device Boot /dev/hdb1 [root@bigboy tmp]# Start 1 End 9729 Blocks 76051710 Id 83 System Linux Putting a Directory Structure on Your New Partition You now need to format the partition.. 63 sectors/track. .. 22 0 22 1 [root@bigboy tmp]# cat /proc/partitions #blocks name -l command 78150744 hdb 78150744 hdb1 tmp]# [root@bigboy tmp]# fdisk -l . to which the new partition will be attached.

You don’t have to wait for a reboot to mount your partition. The fifth entry governs whether the dump filesystem backup command can be used for the filesystem. [root@bigboy tmp]# df -k Filesystem 1K-blocks /dev/hda3 505636 /dev/hda1 101089 none 63028 /dev/hda5 248895 /dev/hda7 3304768 /dev/hda2 3304768 [root@bigboy tmp]# df -k command. which need be only default for most scenarios. If you are not familiar with the /etc/fstab file use the man fstab command to get a full explanation of its various options. The fourth entry defines the mounting options. You can recursively use the command by using the cd command to step down through all the subdirectories until you discover the one with the . You’ll have to add an entry for your new partition that looks like this: # # File: /etc/fstab # /dev/hdb1 /mnt/hdb1 ext3 default 1 2 The first entry is the name of the partition followed by the mount point directory and the filesystem type. The check is done twice. A value of 0 means no. The final entry defines the order in which a filesystem check is done at boot time. Migrating Data to Your New Partition As you remember from investigating with the tion is almost full. it searches the /etc/fstab file for a list of all partitions and their mounting characteristics. and 1 means yes. and then it mounts the partitions automatically. You can use the mount command with the -a option to read the /etc/fstab file for new entries: [root@bigboy tmp]# mount -a You are now able to access your new partition as device /mnt/hdb1. the /var parti- Used Available Use% Mounted on 118224 361307 25% / 14281 81589 15% /boot 0 63028 0% /dev/shm 6613 229432 3% /tmp 2720332 416560 87% /usr 3300536 4232 99% /var The du -sk * command shows the disk usage of all subdirectories in a directory. The root (or master) filesystem has a value of 1 and is checked on the first pass.qxd 2/25/05 10:18 AM Page 440 440 Expanding Disk Capacity Chapter 27 When Linux boots. all other filesystems should have a value of 2.30Harrison_ch27.

. 2... [root@bigboy var]# /var directory to see As a solution. this will later act as a sh-2.30Harrison_ch27.my-web-site..05b# mv /var/transactions /var/transactions-save 5.my-web-site.05b# mkdir /var/transactions 6. Unmount the new /dev/hdb1 partition: sh-2. [root@bigboy tmp]# cd /var [root@bigboy var]# du -sk * 2036 cache 4 db 8 empty ..qxd 2/25/05 10:18 AM Page 441 Adding Disks to Linux 441 greatest file usage. Log into the VGA console.org) 6 12:01 (192-168-1-248. . To migrate the data.. the /var partition will be expanded to the new /dev/hdb1 partition mounted on the /mnt/hdb1 directory mount point. you only had to go to the that the /var/transactions directory was the culprit. Back up the data on the partition you are about to work on. send a message with the wall command informing them that the system is about to shutdown. Copy the contents of the /var/transactions-save directory to the root directory of /dev/hdb1..my-web-site. Create a new. empty mount point: /var/transactions directory. . Rename the /var/transactions directory /var/transactions-save to make sure you have an easy to restore backup of the data.org) 6 16:25 (192-168-1-250..05b# umount /mnt/hdb1 .org) system is shutting down now! 7 15:04:27 2004): Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 3. Use the who command to see who’s logged in. 133784 transactions . In this case. If other users are present. and enter single-user mode: [root@bigboy tmp]# init 1 4. which is actually /mnt/hdb1: sh-2. not just the tapes: sh-2. use these steps: 1.05b# cp -r /var/transactions-save /mnt/hdb1 7. [root@bigboy tmp]# who root pts/0 Nov bob pts/0 Nov bunny pts/0 Nov [root@bigboy tmp]# wall The 6 14:46 (192-168-1-242.

30Harrison_ch27. 11. removing our previous entry for /dev/hdb1 replacing it with one using the new mount point. This exercise showed you how to migrate the entire contents of a subdirectory to a new disk. Test to make sure that the contents of the new /var/transactions directory is identical to /var/transactions-save. # # File: /etc/fstab # #/dev/hdb1 /dev/hdb1 /mnt/hdb1 ext3 default 1 2 ext3 default 1 2 /var/transactions 9. Return to multi-user mode by typing exit. Make sure your applications are working correctly and delete both the /var/transactions-save directory and the /mnt/hdb1 mount point directory at some later date. . you need to become familiar with some LVM concepts: ☞ Physical Volume: A physical volume (PV) is another name for a regular physical disk partition that is used or will be used by LVM.05b# mount -a 10. This is most useful when you find yourself running out of space on a filesystem and want to expand into a new disk partition versus migrating all or a part of the filesystem to a new disk. Remount /dev/hdb1 on the new mount point using the mount -a command. sh-2.qxd 2/25/05 10:18 AM Page 442 442 Expanding Disk Capacity Chapter 27 8. LVM Terms Before you run through the LVM tutorial. Linux also allows you to merge partitions together in order to create a larger combined one.05b# exit 12. EXPANDING PARTITIONS WITH LVM The Logical Volume Manager (LVM) enables you to resize your partitions without having to modify the partition tables on your hard disk. The reasons and steps for doing so will be explained next. which reads /etc/fstab and automatically mounts any entries that are not mounted already: sh-2. Edit the /etc/fstab file. The system will return to its default runlevel.

PEs are important as you usually have to specify the size of your volume group not in gigabytes. has become too full. Unmount Your /home Filesystem As /home stores most users’ data. you’ll need to do some preparatory work before unmounting the filesystem: 1. Let’s go! Configuring LVM Devices It is probably best to learn about the features of LVM through a scenario.org) (192-168-1-250.30Harrison_ch27. You just added a new hard drive /dev/hdf with 50% of the capacity of /dev/hde5 into which you want to expand /home. Another advantage of LVM is that this can all be done without disturbing other partitions on your hard disks. A logical volume is. If other users are present. Take a look at the required steps. The device /dev/hdf has a single partition named /dev/hdf1 into which /dev/hde5 will be merged. The /home filesystem. therefore.my-web-site. ☞ Logical Volumes: Volume groups must then be subdivided into logical volumes.qxd 2/25/05 10:18 AM Page 443 Expanding Partitions with LVM 443 ☞ Volume Group: Any number of physical volumes (PVs) on different disk drives can be lumped together into a volume group (VG).my-web-site. like a virtual partition on your virtual disk drive. ☞ Physical Extent: Real disk partitions are divided into chunks of data called physical extents (PEs) when you add them to a logical volume. but as a number of physical extents.org) is shutting down now! . The LVM process will destroy the data on all physical volumes.org) (192-168-1-248. This may seem complicated. but there isn’t the budget to purchase an adequately sized hard drive. volume groups are analogous to a virtual disk drive. send a message with the wall command informing them that the system is about to shutdown. [root@bigboy tmp]# who root pts/0 Nov 6 14:46 bob pts/0 Nov 6 12:01 bunny pts/0 Nov 6 16:25 [root@bigboy tmp]# wall The system (192-168-1-242. Use the who command to see who’s logged in. Make sure you understand these terms fully as they will be used repeatedly in many of the following sections. which resides on /dev/hde5. Under LVM. Suppose a small company needs to expand disk capacity. Each logical volume can be individually formatted as if it were a regular Linux partition.my-web-site. Back up Your Data Use the tar command or some other method to backup your data in /home. but it allows you to create new virtual partitions with sizes you can change from groups of real disk partitions whose sizes you probably cannot change.

g.05b# fdisk /dev/hde The number of cylinders for this disk is set to 8355. repeat them for /dev/hdf1. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde3 /dev/hde4 /dev/hde5 sh-2.05b# Start 1 4089 5714 6608 6608 End 4088 5713 6607 8355 7500 Blocks 2060320+ 819000 450576 880992 450040+ Id fd 83 83 5 83 System Linux raid autodetect Linux Linux Extended Linux Start fdisk You can change the partition type using fdisk with the disk name as its argument. Determine the Partition Types You have to change each LVM partition used to be of type 8e (Linux LVM). OS/2 FDISK) Command (m for help): . The fdisk examples that follow are for /dev/hde5. which is covered next. DOS FDISK.05b# fdisk -l /dev/hde Disk /dev/hde: 4311 MB. Use it to modify both partitions /dev/hde5 and /dev/hdf1. Log into the VGA console.05b# umount /home Now we’re ready to start modifying the partitions. sh-2.. Unmount the filesystem: sh-2. 63 sectors/track. You can test this with the fdisk -l command. old versions of LILO) 2) booting and partitioning software from other OSs (e.qxd 2/25/05 10:18 AM Page 444 444 Expanding Disk Capacity Chapter 27 Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 7 15:04:27 2004): 2.g. There is nothing wrong with that. and could in certain setups cause problems with: 1) software that runs at boot time (e. but this is larger than 1024.30Harrison_ch27. 4311982080 bytes 16 heads. and enter single-user mode: [root@bigboy tmp]# init 1 3.. Here is an example using /dev/hde that shows your target partitions are of the incorrect type: sh-2.

Command (m for help): t Partition number (1-6): 5 Hex code (type L to list codes): 8e Changed system type of partition 5 to 8e (Linux LVM) Command (m for help): t Partition number (1-6): 6 Hex code (type L to list codes): 8e Changed system type of partition 6 to 8e (Linux LVM) Command (m for help): Make Sure the Change Occurred Use the p command to get the new proposed partition table: Command (m for help): p Disk /dev/hde: 4311 MB. 4311982080 bytes 16 heads. 63 sectors/track. The kernel still uses the old table. 8355 cylinders Units = cylinders of 1008 * 512 = 516096 bytes Device Boot /dev/hde1 /dev/hde2 /dev/hde3 /dev/hde4 /dev/hde5 Start 1 4089 5714 6608 6608 End 4088 5713 6607 8355 7500 Blocks 2060320+ 819000 450576 880992 450040+ Id fd 83 83 5 8e System Linux raid autodetect Linux Linux Extended Linux LVM Command (m for help): Save the Partition Changes Use the w command to permanently save the changes to disk /dev/hde: Command (m for help): w The partition table has been altered! Calling ioctl() to re-read partition table. .qxd 2/25/05 10:18 AM Page 445 Expanding Partitions with LVM 445 Set the ID Type to 8e You now need to set the partition types to the LVM value of 8e. You can also use the L command to get a full listing of ID types in case you forget. WARNING: Re-reading the partition table failed with error 16: Device or resource busy. and then specify the partition number and type code. Partitions /dev/hde5 and /dev/hdf1 are the fifth and sixth partitions on disk /dev/hde. Modify their type using the t command.30Harrison_ch27.

physical volume “/dev/hde5” successfully created sh-2.05b# pvcreate /dev/hdf1 pvcreate -.. use the vgscan command: sh-2. This wipes out all the data on them in preparation for the next step. To do this.05b# vgscan vgscan -.qxd 2/25/05 10:18 AM Page 446 446 Expanding Disk Capacity Chapter 27 The new table will be used at the next reboot.reading all physical volumes (this may take a while. The LVM software effectively tricks the operating system into thinking the volume group is a new hard disk..05b# vgcreate lvm-hde /dev/hdf1 /dev/hde5 Volume group “lvm-hde” successfully created sh-2. sh-2. [root@bigboy updates]# The error above will occur if any of the other partitions on the disk is mounted.05b# pvcreate /dev/hde5 pvcreate -. the volume group is called lvm-hde. This shouldn’t be grave as you are already in single user mode in which most of the system’s processes that would be accessing the partition have been shutdown.) sh-2. Define Each Physical Volume After modifying the partition tables of /dev/hde and /dev/hdf.physical volume “/dev/hdf1” successfully created sh-2. If you haven’t backed up your data yet. initialize the target partitions with the pvcreate command.05b# .05b# Run VGscan The next step is to make Linux scan for any new LVM disk partitions and automatically create the LVM configuration files in the /etc directory. Syncing disks.30Harrison_ch27. do it now! sh-2.05b# Create a Volume Group for the PVs Use the vgcreate command to combine the two physical volumes into a single unit called a volume group. In the example.

You’ll have to know the number of available PEs before creating the logical volume.05b# vgdisplay lvm-hde --. This is done with the vgdisplay command using the new lvm-hde volume group as the argument: sh-2.05b# Format the Logical Volume After the logical volume is created.32 (09-Nov-2002) . You can now use all 212 of them to create a logical volume named lvm0 from volume group lvm-hde.qxd 2/25/05 10:18 AM Page 447 Expanding Partitions with LVM 447 Therefore. 212 PEs are available as free.05b# As you can see. In this case. logical volumes are divided into units called physical extents (PEs).30Harrison_ch27. Like hard disks. the vgcreate syntax uses the name of the volume group as the first argument followed by the partitions that it will be comprised of as all subsequent arguments.05b# lvcreate -l 212 lvm-hde -n lvm0 Logical volume “lvm0” created sh-2. Create a Logical Volume from the Volume Group Now you’re ready to partition the volume group into logical volumes with the lvcreate command. use the -t switch to specify to the mkfs formatting program that you want a type ext3 partition: sh-2. sh-2.99 GB Max PV 256 Cur PV 2 Act PV 2 VG Size 848 MB PE Size 4 MB Total PE 212 Alloc PE / Size 0 / 0 Free PE / Size 212 / 848 MB VG UUID W7bgLB-lAFW-wtKi-wZET-jDJF-8VYD-snUaSZ sh-2.05b# mkfs -t ext3 /dev/lvm-hde/lvm0 mke2fs 1. you can format it as if it were a regular partition. which are divided into blocks of data.Volume group --VG Name lvm-hde VG Access read/write VG Status available/resizable VG # 0 MAX LV 256 Cur LV 0 Open LV 0 MAX LV Size 255.

Now you have to recreate /home on which you’ll later mount your new logical volume: sh-2. This snippet configures the newly labeled partition to be mounted on the /home mount point: /dev/lvm-hde/lvm0 /home ext3 defaults 1 2 /lvm0 The /dev/hde5 and /dev/hdf1 partitions are replaced by the combined logical volume. 98304.qxd 2/25/05 10:18 AM Page 448 448 Expanding Disk Capacity Chapter 27 Filesystem label= OS type: Linux Block size=4096 (log=2) Fragment size=4096 (log=2) 108640 inodes.05b# mkdir /home Update the /etc/fstab File The /etc/fstab file lists all the partitions that need to be automatically mounted when the system boots. Make sure that any reference to them in this file has either been commented with a # character at the beginning of each line or deleted entirely. therefore. you lost the /home directory.00%) reserved for the super user First data block=0 7 block groups 32768 blocks per group. 32768 fragments per group 15520 inodes per group Superblock backups stored on blocks: 32768. whichever comes first.05b# Create a Mount Point When you formatted the /dev/hde5 partition. #/dev/hde5 #/dev/hdf1 /data1 /data2 ext3 ext3 defaults defaults 1 2 1 2 .30Harrison_ch27. don’t want the old partitions to be mounted again. You. Use tune2fs -c or -i to override. sh-2. 217088 blocks 10854 blocks (5. 163840 Writing inode tables: done Creating journal (4096 blocks): done Writing superblocks and filesystem accounting information: done This filesystem will be automatically checked every 38 mounts or 180 days.

It should just contain the lost+found directory. It seems as if when drives get bigger. After mounting. so does the data they are intended to store.05b# mount -a sh-2.05b# ls /home lost+found sh-2. dramatically increasing their speed and capacity too.30Harrison_ch27. but hard disk technology has evolved. CONCLUSION The demise of the hard disk has been predicted for many years. denser memory chips were supposed to eliminate their need. Expanding the existing disk capacity of your server may become an everyday occurrence and the tools described in this chapter should make the task easier. The exit command will make you return to your default runlevel.05b# Restore Your Data You can now restore your backed up data to /home. Faster. They will be around for a long time to come. . Get Out of Single User Mode Return to your original run state by using either the init 3 or init 5 commands. sh-2. test the volume by listing its directory contents.qxd 2/25/05 10:18 AM Page 449 Conclusion 449 Mount the Volume The mount -a command reads the /etc/fstab file and mounts all the devices that haven’t been mounted already.

qxd 2/25/05 10:18 AM Page 450 .30Harrison_ch27.

Enter Single-User Mode As you’ll need to remount the /home filesystem. you may need to restrict the amount of disk space used on each partition by each user or group of users as your disk drives become filled with data. The disk quota feature of Red Hat/Fedora Linux enables you to do this. The procedure is quick: 1. If there are any. Entering single-user mode automatically logs off all users and stops cron jobs. SETTING UP QUOTAS For example. Use the who command to see which users are logged in. besides yourself. send a message stating that the system is about to shutdown with the wall command: 451 /home . you might be able to skip this step. If you are certain that you’re the only user on the system. your family Linux server is running out of space in the filesystem because of an abundance of MP3 downloads. and the setup is fairly simple.31Harrison_ch28.qxd 2/25/05 10:19 AM Page 451 C H A P T E R 28 Managing Disk Usage with Quotas IN THIS CHAPTER ☞ Setting Up Quotas ☞ Other Quota Topics ☞ Conclusion Eventually. you make sure that no other users or processes are using it by first entering single-user mode from the console. so wait until after hours to do this in a business environment.

You can also use the exit command to return to your default runlevel. The old fstab looked like: LABEL=/home /home ext3 defaults 1 2 but your new fstab should be: LABEL=/home /home ext3 defaults.org) 6 16:25 (192-168-1-250. .org) 6 12:01 (192-168-1-248.qxd 2/25/05 10:19 AM Page 452 452 Managing Disk Usage with Quotas Chapter 28 [root@bigboy tmp]# who root pts/0 Nov bob pts/0 Nov bunny pts/0 Nov [root@bigboy tmp]# wall The 6 14:46 (192-168-1-242.05b# mount -o remount /home Get Out of Single-User Mode Return to your original run state by using either the init 3 or init 5 commands. Continue to the next step once the system is back to its normal state. You’ll need to add the usrquota option. Linux needs to reread the file to get its instructions for /home. sh-2. Log into the VGA console. You can do this using the mount command with the -o remount qualifier. eh? Edit Your /etc/fstab File The /etc/fstab file lists all the partitions that need to be auto-mounted when the system boots.my-web-site.my-web-site. In case you forget the name.my-web-site.org) system is shutting down now! 7 15:04:27 2004): Broadcast message from root (pts/0) (Sun Nov The system is shutting down now! [root@bigboy tmp]# 2. and enter single user mode: [root@bigboy tmp]# init 1 Simple. You have to alert Linux that quotas are enabled on the filesystem by editing the /etc/fstab file and modifying the options for the /home directory.31Harrison_ch28. the usrquota option is mentioned in the fstab man pages.usrquota 1 2 Remount the Filesystem Editing the /etc/fstab file isn’t enough.

. You have to generate a quota table. In this case. that lists all the current allocations for each user on the file system. just enable per-user quotas for the /home filesystem: [root@bigboy tmp]# touch /home/aquota.31Harrison_ch28. The man page for quota lists them at the bottom.user file (defines quotas by user). or both.Quotafile /home/aquota. The edquota command enables you to selectively edit a portion of the aquota. Linux compares the values in this table with the quota limitations that the systems administrator has placed in the aquota files and uses this information to determine whether the user has rights to increased disk usage.user [root@bigboy tmp]# Initialize the Quota Table Editing the /etc/fstab file and remounting the file system only alerted Linux to the fact that the filesystem has quota capabilities.group file (defines quotas by group).user was probably truncated.qxd 2/25/05 10:19 AM Page 453 Setting Up Quotas 453 Create the Partition Quota Configuration Files The uppermost directory of the filesystem needs to have an aquota.user file on a per-user basis: [root@bigboy tmp]# edquota -u mp3user The command invokes the vi editor.. This table will be automatically and transparently updated each time a file is modified. separate from the aquota files. You initialize the table with the quotacheck command. Disk quotas for user mp3user (uid 503): Filesystem blocks soft hard /dev/hda3 24 0 0 inodes 7 soft 0 hard 0 . an aquota. quotacheck: Scanning /dev/hda3 [/home] done quotacheck: Checked 185 directories and 926 files [root@bigboy tmp]# Edit the User’s Quota Information Now you need to edit the user’s quota information.user [root@bigboy tmp]# chmod 600 /home/aquota. Be prepared: You’ll get an error the first time you enter the command. Can’t save quota settings. because Linux will realize that the aquota file wasn’t created using one of the quota commands: [root@bigboy tmp]# quotacheck -vagum quotacheck: WARNING .

the user is barred from using additional disk space or files. in 1K blocks. it invokes the vi editor. Like the The grace period is a time limit before the soft limit is enforced for a quota-enabled file system. You can use time units of seconds. weeks. You’ll need to make Linux scan its hard disks periodically to check for exceeded quotas. and months. ☞ Soft limit: The maximum blocks/inodes a quota user may have on a partition. that the user is currently using. With grace periods. minutes. When the grace period expires. but they can never exceed their hard limit. limits are disabled. When set to zero. you can edit a number of fields: ☞ Blocks: The amount of space. ☞ Hard limit: The maximum blocks/inodes a quota user may have on a partition when a grace period is set. If the values are exceeded. The for each filesystem. days. ☞ Inodes: The number of files the user is currently using. Users may exceed a soft limit.31Harrison_ch28. OTHER QUOTA TOPICS Creating disk quotas frequently isn’t enough. The role of a soft limit changes if you use grace periods. hours. then Linux prevents the creation of new files or the expansion of existing files to use more disk space. Editing Grace Periods edquota -t command sets the grace period edquota -u command. This is what you’ll see with the command edquota -t: . You also have to manage the process by reviewing the quota needs of each user and adjusting them according to the policies of your company. the user is warned only that their soft limit has been exceeded. This section describes the most common quota management activities.qxd 2/25/05 10:19 AM Page 454 454 Managing Disk Usage with Quotas Chapter 28 From here. Here user mp3user is limited to a maximum of 5MB of data storage on /dev/hda3 (/home): Disk quotas for user mp3user (uid 503): Filesystem blocks soft hard /dev/hda3 24 5000 0 inodes 7 soft 0 hard 0 Testing Linux checks the total amount of disk space a user uses each time a file is accessed and compares it against the values in the quota file. depending on the configuration.

qxd 2/25/05 10:19 AM Page 455 Other Quota Topics 455 [root@bigboy tmp]# edquota -t Grace period before enforcing soft limits for users: Time units may be: days. “Expanding Disk Capacity. but also reduces the likelihood of having fragmented disk structures that slow disk access due to the presence of too many files... hours. Chapter 27.” covers a variety of approaches to increasing your storage capacity. Getting Quota Reports The repquota command lists quota usage limits of all users on the system: [root@bigboy tmp]# repquota /home *** Report for user quotas on device /dev/hda3 Block grace time: 7days. . but 7 days is wrong... It helps to not only limit the use of physical disk resources that could potentially be put to better use. users must be made aware of the challenges you face as a result of excessive disk usage. there comes a time when you are faced with no option but to expand the number of disks in your system. or seconds Filesystem Block grace period Inode grace period /dev/hda3 7days 7days Note There should be no spaces between the number and the unit of time measure. 7days is correct. Of course... mp3user -24 0 0 7 0 0 [root@bigboy tmp]# -g CONCLUSION Disk quotas are good to put in place for specific users whose disk usage activities need to be curtailed. Therefore. minutes. User education must play a major role in your strategy too. Editing Group Quotas Editing quotas on a per-group basis can be done similarly with the edquota command. . Inode grace time: 7days Block limits File limits User used soft hard grace used soft hard grace ---------------------------------------------------------------------root -52696 0 0 1015 0 0 .31Harrison_ch28. .

31Harrison_ch28.qxd 2/25/05 10:19 AM Page 456 .

the directories of their filesystems to a directory of an already existing filesystem. This in effect makes the new hard disk transparently appear to be a subdirectory of the filesystem to which it is attached. Basic configuration is fairly simple. NFS OPERATION OVERVIEW Linux data storage disks contain files stored in filesystems with a standardized directory structure. New disks are added by attaching. NFS was developed to allow a computer system to access directories on remote computers by mounting them on a local filesystem as if they were a local disk.32Harrison_ch29. and this chapter will explain all the essential steps. The systems administrator on the NFS server has to define the 457 . The Network File System protocol (NFS) is used when disks need to be shared between Linux servers.qxd 2/25/05 10:19 AM Page 457 C H A P T E R 29 Remote Disk Access With NFS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ NFS Operation Overview Installing nfs Scenario Configuring NFS on the Server Configuring NFS on the Client Activating Modifications to the /etc/exports File The NFS automounter Troubleshooting NFS Other NFS Considerations Conclusion Samba is usually the solution of choice when you want to share disk space between Linux and Windows machines. or mounting.

for access by the NFS clients. or obscured. With NFS. when a server fails. Key NFS Concepts Data access over a network always introduces a variety of challenges. ☞ Only export local filesystems. in favor of the files in the mounted filesystem. Normally. as in the case of NFS. VFS also translates these requests to match the actual filesystem format on the NFS server’s disks. then the original files in the directory reappear unchanged. file accesses timeout and the file pointers are reset to zero.32Harrison_ch29. ☞ Do not export a subdirectory of a directory that has already been exported—the exception being when the subdirectory is on a different physical device. General NFS Rules You should follow some general rules when configuring NFS: ☞ Only export directories beneath the / directory. This is done in such a way that files on the remote NFS server appear to the user to be no different than those on a local disk. Keep in mind that when you mount any filesystem on a directory. When the filesystem is unmounted. and administrators on the clients need to define both the NFS server and the subset of its exported directories to use. the original contents of the directory are ignored. VFS The virtual filesystem (VFS) interface is the mechanism used by NFS to transparently and automatically redirect all access to NFS-mounted files to the remote server. Likewise. Here are some key NFS background concepts that will help in your overall understanding. This allows the NFS server to be not only a completely different operating system. Stateless Operation Programs that read and write to files on a local filesystem rely on the operating system to track their access location within the file with a pointer. or exported. it was decided that the NFS client daemon would act as a failsafe intermediary between regular programs running on the NFS client and the NFS server. and networks can be unreliable. especially if the operation is intended to be transparent to the user. but also use different naming schemes for files with different file attribute types. As NFS is a network-based file system. do not export the parent of a subdirectory unless it is on a separate device.qxd 2/25/05 10:19 AM Page 458 458 Remote Disk Access With NFS Chapter 29 directories that need to be activated. the NFS server doesn’t maintain the file pointer .

32Harrison_ch29. which is approximately a week. then the client retries the mount process until the NFS number of retries has been exceeded. is called a hard mount. Also. then an error will occur. The NFS server caches information too. whether in the background or foreground. If there is a link to the . With soft mounts. NFS attempts to guarantee the consistency of your data with these constant retries. The default is 10. Caching NFS clients typically request more data than they need and cache the results in memory locally so that further sequential access of the data can be done locally versus over the network. This means that if an NFS server suddenly fails.qxd 2/25/05 10:19 AM Page 459 NFS Operation Overview 459 information. relative symbolic links are interpreted relative to the client’s filesystem.. You’ll have to be careful not to obscure the contents of this original directory in the process. Consider an example where the /data1 directory on the server is mounted on the /data1 directory.000 minutes. Unlike absolute links./data2 directory on the NFS server and a directory corresponding to . the NFS client does. Caching therefore helps to reduce the amount of network traffic while simultaneously improving the speed of some types of data access. such as the directory information for the most recently accessed files and a read ahead cache for recently read files.. It is possible to use a bg option to spawn the retries off as a subprocess so that the main mount command can continue to process other requests. Plan carefully before doing this. the NFS client can precisely restart the file access once more after patiently waiting until the server returns online. If an absolute link points to a directory on the NFS server that hasn’t been exported. repeated RPC failures . This is also known as a read ahead cache. If the mount cannot occur during the default RPC timeout period. mounting a filesystem on a symbolic link actually mounts the filesystem on the target of the symbolic link. Data that’s to be written to the NFS server is cached with the data being written to the server when the cache becomes full. The difficulty here is that if the NFS server is unavailable. the mount command will hang for a week until it returns online./data2 doesn’t exist on the NFS client. then the NFS client won’t be able to access it. NFS and Symbolic Links You have to be careful with the use of symbolic links on exported NFS directories. Hard and Soft Mounts The process of continuous retrying. NFS Background Mounting NFS clients use the remote procedure call (RPC) suite of network application helper programs to mount remote filesystems.

usually above portmap: . well known TCP port. This chapter focuses on version 2. but with the additions that: ☞ File locking and mounting are integrated in the NFS daemon and operate on a single. The advantage is that the operation completes quickly. whether it fails or not. This is then used to negotiate a range of TCP ports. 3. portmap manages connections for applications that use the RPC specification.32Harrison_ch29. whereas before it was optional ☞ Support for the bundling of requests from each client provides more efficient processing by the NFS server It is important to match the versions of NFS running on clients and servers to help ensure the necessary compatibility to get NFS to work predictably. NFS Versions Three versions of NFS are available currently: versions 2. Version 1 was a prototype. making network security easier ☞ File locking is mandatory. if this is the case it is best not to place critical data that needs to be updated regularly or executable programs in such a location. but a suite of interrelated programs that work together to get the job done: ☞ The primary daemon upon which all the others rely. Important NFS Daemons NFS isn’t a single program. By default.1 bytes ☞ Supports the NFS server data updates as being successful when the data is written to the server’s cache ☞ Negotiates the data limit per read or write request between the client and server to a mutually decided optimal value Version 4 maintains many of version 3’s features. and 4.qxd 2/25/05 10:19 AM Page 460 460 Remote Disk Access With NFS Chapter 29 cause the NFS operation to fail not hang and data consistency is therefore not guaranteed. portmap listens to TCP port 111 on which an initial connection is made. The disadvantage is that the use of the soft option implies that you are using an unreliable NFS server. which: ☞ Supports files up to 4GB long ☞ Requires an NFS server to successfully write data to its disks before the write request is considered successful ☞ Has a limit of 8KB per read or write request The main differences with version 3 are that it: ☞ Supports extremely large file sizes of up to 264 .

When you use rpm in conjunction with grep.0. .qxd 2/25/05 10:19 AM Page 461 Installing nfs 461 port 1024. see Chapter 6. The nfslock daemon needs to be run on both the NFS server and client. ☞ netfs: Allows RPC processes run on NFS clients to mount NFS filesystems on the server.0-57 [root@bigboy tmp]# A blank list means that you’ll have to install the required packages. Now take a look at how to configure these daemons to create functional NFS client/server peering. “Installing RPM Software.1. to be used for subsequent data transfers.1. they can be added fairly easily once you find the nfs-utils and portmap RPMs.rpm. If nfs and portmap are not installed.3-1 nfs-utils-1. You also need to have the RPC portmap package installed. The nfs daemon needs to be run on the NFS server only. and the rpm command can tell you whether it’s on your system already. The nfslock daemon needs to be run on the NFS client only.”) Remember that RPM filenames usually start with the software’s name and a version number. as in nfs-utils-1. you can determine all the portmap applications installed: [root@bigboy tmp]# rpm -q portmap portmap-4.32Harrison_ch29. ☞ nfslock: Used to allow NFS clients to lock files on the server via RPC processes.1-3. You need to run portmap on both the NFS server and client.i386.9 [root@bigboy tmp]# A blank list means that you’ll have to install the required packages. ☞ nfs: Starts the RPC processes needed to serve shared NFS file systems. You can determine whether you have nfs installed using the RPM command in conjunction with the grep command to search for all installed nfs packages: [root@bigboy tmp]# rpm -qa | grep nfs redhat-config-nfs-1.31. INSTALLING NFS Red Hat Linux installs nfs by default. and also by default nfs is activated when the system boots. (If you need a refresher.

#/etc/exports /data/files /home /data/test /data/database *(ro. and the second part lists NFS options in brackets. as well as a correctly configured /etc/exports file.my-web-site. nfs.0 /24 network. Another Linux server on the network has additional disk capacity in its /data partition and the office would like to expand into it as an interim expansion NFS server. The office cannot tolerate any down time.203/32(rw. because the server is accessed by overseas programmers and clients at night and local ones by day.sync) 192.1. use the sync option to ensure that file data cached in memory is automatically written to the disk after the completion of any disk data copying operation.168.org(rw. you need to activate the settings.255 ☞ Read/write access to the /data/test directory from servers in the my-web-site.0 to 192.0/24(rw. CONFIGURING NFS ON THE SERVER Both the NFS server and NFS client have to have parts of the NFS package installed and running. even after hours.168. The second column has two parts.1. and it consists of two columns.1. and nfslock operational.sync) After configuring your /etc/exports file.1. which is all addresses from 192. 192.168. Budgets are tight and the company needs a quick solution until it can get a purchase order approved for a hardware upgrade. . but first make sure that NFS is running correctly.sync) *. The /etc/exports File The /etc/exports file is the main NFS configuration file.1. The first part lists the networks or DNS domains that can get access to the directory.203 In all cases.org DNS domain ☞ Read/write access to the /data/database directory from a single server.168.32Harrison_ch29. The server needs portmap.1.sync) 192.qxd 2/25/05 10:19 AM Page 462 462 Remote Disk Access With NFS Chapter 29 SCENARIO A small office has an old Linux server that is running out of disk space. The first column lists the directories you want to make available to the network.168. For the scenario you need: ☞ Read-only access to the /data/files directory to all networks ☞ Read/write access to the /home directory from all servers on the 192. Here’s how to do it.168.

You should get a listing of running RPC programs that must include mountd.qxd 2/25/05 10:19 AM Page 463 Configuring NFS on the Client 463 Starting NFS on the Server Configuring an NFS server is straightforward: 1. nfs. . [root@bigboy tmp]# chkconfig --level 35 nfs on [root@bigboy tmp]# chkconfig --level 35 nfslock on [root@bigboy tmp]# chkconfig --level 35 portmap on 2. Use the chkconfig command to configure the required nfs and RPC portmap daemons to start at boot. or mount point. and nlockmgr: [root@bigboy tmp]# rpcinfo -p localhost program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100003 2 udp 2049 nfs 100003 3 udp 2049 nfs 100021 1 udp 1024 nlockmgr 100021 3 udp 1024 nlockmgr 100021 4 udp 1024 nlockmgr 100005 1 udp 1042 mountd 100005 1 tcp 2342 mountd 100005 2 udp 1042 mountd 100005 2 tcp 2342 mountd 100005 3 udp 1042 mountd 100005 3 tcp 2342 mountd [root@bigboy tmp]# CONFIGURING NFS ON THE CLIENT NFS configuration on the client requires you to start the NFS application.d directory to start the nfs and RPC portmap daemons. and finally to mount the NFS server’s directory on your local directory. You also should activate NFS file locking to reduce the risk of corrupted data. Test whether NFS is running correctly with the rpcinfo command.32Harrison_ch29. but when needed. Here’s how to do it all. portmapper. you can also stop and restart the processes with the stop and restart options: [root@bigboy tmp]# service portmap start [root@bigboy tmp]# service nfs start [root@bigboy tmp]# service nfslock start 3. create a directory on which to mount the NFS server’s directories that you exported via the /etc/exports file. Use the init scripts in the /etc/init. The examples use the start option.

my-website.arpa domain name pointer 192-168-1-102. . portmapper. As on the server.in-addr. In other words. Use the chkconfig command to configure the required nfs and RPC portmap daemons to start at boot. a DNS lookup on the NFS server for the IP address of the NFS client must return a server name that will map back to the original IP address when a DNS lookup is done on that same server name. Test whether NFS is running correctly with the rpcinfo command.168.d directory to start the nfs and RPC portmap daemons. and nlockmgr: [root@smallfry root]# rpcinfo -p program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100024 1 udp 32768 status 100024 1 tcp 32768 status 100021 1 udp 32769 nlockmgr 100021 3 udp 32769 nlockmgr 100021 4 udp 32769 nlockmgr 100021 1 tcp 32769 nlockmgr 100021 3 tcp 32769 nlockmgr 100021 4 tcp 32769 nlockmgr 391002 2 tcp 32770 sgi_fam [root@smallfry root]# NFS and DNS The NFS client must have a matching pair of forward and reverse DNS entries on the DNS server used by the NFS server. The listing of running RPC programs you get must include status.qxd 2/25/05 10:19 AM Page 464 464 Remote Disk Access With NFS Chapter 29 Starting NFS on the Client Three more steps easily configure NFS on the client: 1.192. but you can also stop and restart the processes with the stop and restart options: [root@smallfry tmp]# service portmap start [root@smallfry tmp]# service netfs start [root@smallfry tmp]# service nfslock start 3.1.1.32Harrison_ch29. the examples use the start option. [root@bigboy tmp]# host 192.168. Use the init scripts in the /etc/init.102 201.org. Activate nfslock to lock the files and reduce the risk of corrupted data: [root@smallfry tmp]# chkconfig --level 35 netfs on [root@smallfry tmp]# chkconfig --level 35 nfslock on [root@smallfry tmp]# chkconfig --level 35 portmap on 2.

my-website. you need to edit the /etc/fstab file if you need the NFS directory to be made permanently available to users on the NFS.nfsvers=2 This example used the soft and nfsvers options. Table 29.1 Option bg fg soft hard Possible NFS Mount Options Description Retry mounting in the background if mounting initially fails.forward lookup doesn’t exist Making NFS Mounting Permanent In most cases.mountd: Fake hostname smallfry. Use soft mounting.102 [root@bigboy tmp]# This is a security precaution added into the nfs package that lessens the likelihood of unauthorized servers from gaining access to files on the NFS server.org 192-168-1-102.32Harrison_ch29. This requires an entry in the /etc/fstab file in addition to the creation of the mount point directory.100:/data/files 0 Mount Point /mnt/nfs Type nfs Options Dump 0 soft.org for 192. Table 29.1 .1 outlines these and other useful NFS mounting options you may want to use.1.1. Mount in the foreground.168.168.my-web-site. Use hard mounting.168. continues . Therefore.org has address 192.my-web-site.100) as an NFS-type filesystem using the local /mnt/nfs mount point directory: #/etc/fstab #Directory FSCK 192.qxd 2/25/05 10:19 AM Page 465 Configuring NFS on the Client 465 [root@bigboy tmp]# host 192-168-1-102. mount the /data/files directory on server Bigboy (IP address 192.1. Failure to correctly register your server IPs in DNS can result in “fake hostname” errors: Nov 7 19:14:40 bigboy rpc. For the example.16801. See the NFS man pages for more details. The /etc/fstab File The /etc/fstab file lists all the partitions that need to be auto-mounted when the system boots. users want their NFS directories to be permanently mounted.

The default is dependent on the kernel. The version of NFS the mount command should attempt to use. as you’ll see. set it to 8192 to assure maximum throughput.100’ failed: server is down. To be safe. Failure to do so may result in an error message: [root@probe-001 tmp]# mount -a mount to NFS server ‘192. The amount of data NFS will attempt to access per write operation.qxd 2/25/05 10:19 AM Page 466 466 Remote Disk Access With NFS Chapter 29 Table 29.168. which most NFS clients use. Note There are multiple versions of NFS. Permanently Mounting the NFS Directory You’ll now create a mount point directory. /mnt/nfs.1. the default is UDP. thereby making this a permanent NFS mount.1 Option rsize=n continued Description The amount of data NFS will attempt to access per read operation. allow for the process to be aborted using the usual methods such as CTRL-C and the kill command. it reads the /etc/fstab file and executes the command. Attempt to mount the filesystem using TCP packets. this changes after the mounting is completed: [root@smallfry tmp]# mkdir /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs [root@smallfry tmp]# mount -a [root@smallfry tmp]# ls /mnt/nfs ISO ISO-RedHat kickstart RedHat [root@smallfry tmp]# mount -a Each time your system boots. For NFS version 2. The default is dependent on the kernel. Newer NFS servers may also be able to handle NFS version 4. If the filesystem is hard mounted and the mount times out. on which to mount the remote NFS directory and then use the mount -a command to activate the mount. For NFS version 2.32Harrison_ch29. wsize=n nfsvers=n tcp intr The steps to mount the directory are fairly simple. [root@probe-001 tmp]# . it is best to force the NFS server to export directories as version 2 using the nfsvers=2 option in the /etc/fstab file as shown in the example. set it to 8192 to assure maximum throughput. Notice how before mounting there were no files visible in the /mnt/nfs directory. the most popular of which is version 2.

” In this case. this changes after the mounting is complete: [root@smallfry tmp]# mkdir /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs [root@smallfry tmp]# mount -t nfs 192. Notice how before mounting there were no files visible in the /mnt/nfs directory. you’re mounting the /data/files directory as an NFS-type filesystem on the /mnt/nfs mount point.32Harrison_ch29.100. This is a manual process. New Exports File When no directories have yet been exported to NFS. you can use the export only the new entries: [root@bigboy tmp]# exportfs -r exportfs -r command to . ACTIVATING MODIFICATIONS TO THE /ETC/EXPORTS FILE You can force your system to re-read the /etc/exports file by restarting NFS. In a nonproduction environment. The NFS server is Bigboy whose IP address is 192.168. this may cause disruptions when an exported directory suddenly disappears without prior notification to users.qxd 2/25/05 10:19 AM Page 467 Activating Modifications to the /etc/exports File 467 Manually Mounting NFS File Systems If you don’t want a permanent NFS mount.100:/data/files /mnt/nfs [root@smallfry tmp]# ls /mnt/nfs ISO ISO-RedHat kickstart RedHat [root@smallfry tmp]# Congratulations! You’ve made your first steps towards being an NFS administrator.1. use the command: [root@bigboy tmp]# exportfs -a exportfs -a Adding a Shared Directory To An Existing Exports File When adding a shared directory.168. then you can use the mount command without the /etc/fstab entry to gain access only when necessary. Here are some methods you can use to update and activate the file with the least amount of inconvenience to others.1. see the section “The NFS automounter. for an automated process.

after the timeout period the remote directory is unmounted and the mount point is deleted. The steps are: 1. The mount point directory is dynamically created by the automounter when needed. When the kernel makes an NFS file request. THE NFS AUTOMOUNTER The permanent mounting of filesystems has its disadvantages. Comment out the corresponding entry in the NFS server’s file and reload the modified file: [root@bigboy tmp]# exportfs -ua [root@bigboy tmp]# exportfs -a /etc/exports You have now completed a seamless removal of the exported directory with much less chance of having critical errors. .32Harrison_ch29. However. after which time it is unmounted automatically. becomes more difficult. instead using an NFS-specific map file that can be distributed to multiple clients. You can add mounting options to a third column. In addition. the mount is permanent. Moving. you can use the file to specify the expected duration of the NFS mount. Also. Unexport the mount point directory on the NFS client using the umount command. automounter intercepts it and mounts the remote directory on the mount point defined in the map file. automounter Map Files The master map file of automounter has a simple format that defines the name of the mount point directory in the first column and the subsidiary map file that controls its mounting in the second. 2. NFS uses an automounter feature that overcomes these shortcomings by allowing you to bypass the /etc/fstab file for NFS mounts. NFS client management. the /etc/fstab file is unique per Linux server and has to be individually edited on each. therefore. automounter continues to report to the operating system kernel that the mount is still active. you’re unmounting the /mnt/nfs mount point: [root@smallfry tmp]# umount /mnt/nfs Note You may also need to edit the /etc/fstab file of any entries related to the mount point if you want to make the change permanent even after rebooting.qxd 2/25/05 10:19 AM Page 468 468 Remote Disk Access With NFS Chapter 29 Deleting. In this case. tying up system resources even when the NFS server isn’t being accessed. or Modifying a Share Removing an exported directory from the /etc/exports file requires work on both the NFS client and server. For example.

Finally. an early release of Fedora Core 3 doesn’t appear to work correctly with automounter. If one does. then automounter mounts the files for the system. If it finds one. Note Based on preliminary testing.master file. A good example would be all the users’ directories under /home. No errors occur in the logs either. # # File: /etc/auto. then automounter checks the subsidiary map file for a subdirectory mount point key. it refers to the keys in the /etc/auto. the maps don’t appear to be activated.32Harrison_ch29. . You have to have one indirect map defined to avoid startup errors.master file. and the third column lists the NFS servers and the filesystems that map to the keys.qxd 2/25/05 10:19 AM Page 469 The NFS automounter 469 In the example. It is also the name of the mount point under the directory listed in the /etc/auto. The second column provides all the NFS options to be used.master file to see whether any fall within the realm of the automounter’s responsibility.master # /home /etc/auto. except that columns two and three have been switched. the mount will only last for five minutes (300 seconds). The Structure of Direct and Indirect Map Files The format of these map files is similar to that of the /etc/auto. and this value will act as a default for all the entries in the /etc/auto.direct file.direct Direct Maps Direct maps are used to define NFS filesystems that are mounted on different servers or that don’t all start with the same prefix. and after doing so.home --timeout=300 //etc/auto. the /home directory needs to be automounted on an NFS server and the configuration information is defined in the /etc/auto. Irregular entries that don’t match /home are placed in the /etc/auto. Indirect Maps Indirect maps define directories that can be mounted under the same mount point. Column one lists all the directory keys that will activate the automounter feature.home file. When the NFS client accesses a file.home file.

This means all direct mount points in the same directory are mounted simultaneously even if only one of them is being accessed.home bigboy:/home/peter ochorios:/home/bob waitabit:/home/bunny Direct Map File Example The second entry in the /etc/auto. # # File: # peter bob bunny /etc/auto.soft bigboy:/disk1/data/sales waitabit:/var/mysql/database Note The automounter treats direct mounts as if they were files in a directory. if you don’t then you’ll get an error like this in your /var/log/messages file: Nov 7 19:24:12 smallfry automount[31801]: bad map format: found indirect. and bunny. and Waitabit. and you can use absolute path names with direct map files only.home file. bob.32Harrison_ch29. perhaps the simplest is just to place direct mount points in different directories. # # File: /etc/auto.master file redirected all references to the /home directory to the /etc/auto.master file must all begin with /-. Ochorios.qxd 2/25/05 10:19 AM Page 470 470 Remote Disk Access With NFS Chapter 29 Indirect Map File Example In the previous example. not as individual directories. Note Direct map entries in the /etc/auto. the /etc/auto.direct # /data/sales -rw /sql/database -ro. these directories are actually mount points for directories on servers Bigboy. This second file has entries for peter. In the example the /data/sales and /sql/database are the mount points for directories on servers Bigboy and Waitabit.master file was specifically created to handle all references to one of a kind directory prefixes. There are tricks you can use to avoid this. This can cause excessive mounting activity that can slow response times. expected direct exiting .

If the access was for /home/peter.home # * bigboy:/home/& Starting automounter Fedora Linux installs the automounter RPM. If access was for /home/bob. # # File: /etc/auto. which means all. which instructs automounter to substitute the value of the key for the & character.home # peter bigboy:/home/& Using the Asterisk Wildcard In the example below. Use the chkconfig command to configure the automounter daemons to start at boot. the key is *. the key is peter. then the ampersand is interpreted to mean peter. then bigboy:/home/bob would be mounted.32Harrison_ch29.d directory to start the automounter daemons. Using the Ampersand Wildcard In the example below. Activate NFS file locking to reduce the risk of corrupted data. so the ampersand wildcard is interpreted to mean peter too. and the ampersand (&). Here are some quick steps to get automounter started: 1. Use the init scripts in the /etc/init. The example uses the start option.qxd 2/25/05 10:19 AM Page 471 The NFS automounter 471 Wildcards in Map Files You can use two types of wildcards in a map file. This means you’ll be mounting the bigboy:/home/peter directory. but you can also stop and restart the process with the stop and restart options. # # File: /etc/auto. by default. called autofs. [root@bigboy tmp]# service autofs start .home file. meaning that automounter will attempt to mount any attempt to enter the /home directory. [root@bigboy tmp]# chkconfig autofs on 2. But what’s the value of the ampersand? It is actually assigned the value of the key that triggered the access to the /etc/auto. and bigboy:/home/peter is mounted. The asterisk (*).

The most common problems are caused by forgetting to start NFS. Typographical error in the name of the server. Table 29.qxd 2/25/05 10:19 AM Page 472 472 Remote Disk Access With NFS Chapter 29 3. but you’ll also have to make sure basic connectivity has been taken care of as outlined in Chapter 4. and accounting control.” contains a full scenario in which a school computer laboratory uses automounter to centrally house all the home directories of its students. Option Too many levels of remote in path Permission denied No such host No such file or directory . “Centralized Logins Using NIS. Use the pgrep command to determine whether automounter is running. automounter Examples Now that you understand NFS automounter. no troubleshooting plan would be complete without frequent reference to the /var/log/messages file when searching for additional clues. This could be the client’s root user who has unprivileged status on the server due to the root_squash option. If it is. [root@bigboy tmp]# pgrep automount 32261 [root@bigboy tmp]# As you can see. Table 29. Typographical error in the name of the file or directory. TROUBLESHOOTING NFS A basic NFS configuration usually works without problems when the client and server are on the same network.” Sometimes a firewall being present on the path between the client and server can cause difficulties. you may benefit from an example. “Simple Network Troubleshooting. they don’t exist. these checks still apply. User is denied access. Another common cause of failure is the iptables firewall daemon running on either the server or client without the administrator realizing it. to edit the /etc/fstab file.2 Some Common NFS Error Messages Description Attempting to mount a filesystem that has already been mounted. managing the startup of automounter is very similar to that of other Linux applications and should be easy to remember. When the client and server are on different networks.32Harrison_ch29.2 shows some common NFS errors you’ll encounter. Could also be because the user on the client doesn’t exist on the server. Chapter 30. the command will return the process ID of the automount daemon. Additional centralization is also achieved by using NIS for login authentication. access. or to export the /etc/exports file. As always.

168. the root_squash mount option will prevent the root user from doing this. [nfsuser@smallfry nfsuser]$ df -F nfs Filesystem 1K-blocks Used Available Use% Mounted on 192. In many cases. in this case one client has an IP address of 192. Option NFS server is not responding Stale file handle Fake hostname The showmount Command When run on the server.3. Run it on the NFS client to verify that NFS mounting has occurred.32Harrison_ch29. while the -c option provides them for clients. the showmount -a command lists all the currently exported directories. A file that was previously accessed by the client was deleted on the server before the client closed it.168.1.1.2 Some Common NFS Error Messages Description The server could be overloaded or down. so it’s best to try it as an unprivileged user.102:* [root@bigboy tmp]# The df Command The df command lists the disk usage of a mounted filesystem. The -s option provides NFS server stats.168.1.qxd 2/25/05 10:19 AM Page 473 Troubleshooting NFS 473 Table 29.102: [root@bigboy tmp]# showmount -a All mount points on bigboy: *:/home 192. Threshold guidelines are provided in Table 29. Forward and reverse DNS entries don’t exist for the NFS client.100:/home/nfsuser 1032056 346552 633068 36% /home/nfsuser [nfsuser@smallfry nfsuser]$ The nfsstat Command The nfsstat command provides useful error statistics. It also shows a list of NFS clients accessing the server. [root@bigboy tmp]# nfsstat -s Server rpc stats: calls badcalls badauth 1547 0 0 Server nfs v2: null getattr setattr badclnt 0 root xdrcall 0 lookup readlink .

Bad RPC requests.qxd 2/25/05 10:19 AM Page 474 474 Remote Disk Access With NFS Chapter 29 244 read 0 link 0 100% 0 0% 0 wrcache write 0% 0 0% 0 symlink mkdir 0% 0 0% 0 0% 0 create 0% 0 rmdir 0% 0 0% 0 0% 0 remove rename 0% 0 0% 0 readdir fsstat 0% 0 0% 0 0% 0% 0% Server nfs v3: null getattr setattr 251 19% 332 25% 0 0% read write create 39 2% 14 1% 1 0% remove rmdir rename 0 0% 0 0% 0 0% fsstat fsinfo pathconf 1 0% 34 2% 0 0% [root@bigboy tmp]# lookup access readlink 265 20% 320 24% 0 0% mkdir symlink mknod 1 0% 0 0% 0 0% link readdir readdirplus 0 0% 0 0% 31 2% commit 14 1% Table 29.3 Value Readlink Error Thresholds for the nfsstat Command Threshold > 10% Description Excessive numbers of symbolic links slowing performance. Check the noac and wsize mount options. This value tracks the percentage of file attribute reads that are not from cache refresh requests. which prevents file attribute caching. Getattr > 50% badcalls >0 retrans > 5% Writes Read > 10% > 10% OTHER NFS CONSIDERATIONS NFS can be temperamental. Often these issues can be resolved with some basic guidelines outlined in this section. Usually caused by the NFS noac mount option. and you have to be aware of the file permissions on both the NFS client and server to get it to work correctly. is cached in NFS.32Harrison_ch29. An incorrect configuration can cause it to be unresponsive. or having a user in too many groups. . the root user attempting to access data governed by the root_squash directive. File attributes. Check the noac and rsize mount options. Percentage of requests for service that the client had to retransmit to the servers. Could be due to slow NFS servers or poor network conditions. Writes are slow due to poor caching values. Could be due to poorly configured authentication. Try to replace them with a directory and mount the filesystem directly on this new mount point. Reads are slow due to poor caching values. like file data. Its security is relatively weak.

As a result. ☞ Do not include NFS-mounted directories as part of your search path. ☞ Export only the most needed data.qxd 2/25/05 10:19 AM Page 475 Other NFS Considerations 475 Security NFS and portmap have had a number of known security deficiencies in the past. the NFS client waits indefinitely for it to return. unless the server’s reliability can be guaranteed. . This also forces programs relying on the same client server relationship to wait indefinitely too. ☞ Use the root_squash option in /etc/exports file (default) to reduce the risk of the possibility of a root user on the NFS client having root file permission access on the NFS server. You can exercise security-related caution with NFS by following a few guidelines: ☞ Restrict its use to secure networks. I’d suggest that you refer to a dedicated NFS reference for more detailed advice. This causes NFS to report an I/O error to the calling program after a long timeout. These points should be the foundation of your NFS security policy. I don’t recommended using NFS over insecure networks. if the NFS server fails. because a hung NFS connection to a directory in your search path could cause your shell to pause at that point in the search path until the NFS session is regained. isn’t comprehensive due to the concise scope of this book.32Harrison_ch29. ☞ Avoid having NFS servers that NFS mount each other’s filesystems or directories. ☞ Consider using read-only exports whenever data updates aren’t necessary. NFS Hanging As stated before. For this reason. This is normally an undesirable condition. ☞ Always use the sync option whenever possible. especially if the NFS client and NFS server are being managed by different sets of administrators. the list. use the soft option in the NFS client’s /etc/fstab file. NFS doesn’t encrypt data and it is possible for root users on NFS clients to have root access to the server’s filesystems. You can reduce the risk of NFS hanging by taking a number of precautions: ☞ Run NFS on a reliable network. ☞ Do not have mission-critical computers rely on an NFS server to operate. however.

but NFS has a history of not handling file locking well. therefore. File Permissions The NFS file permissions on the NFS server are inherited by the client. . intermittent or insecure. especially if the NFS client and NFS server are being managed by different sets of administrators. If a user named nfsuser exists on the NFS client. allowing you more time to focus on other important administrative tasks. A main communication control channel is usually created between the client and server on TCP port 111. This can be disabled with the no_root_squash export option in the /etc/exports file. Test your network-based applications thoroughly before considering using NFS. Restricting Access to the NFS Server NFS doesn’t provide restrictions on a per-user basis. There are ways to limit the TCP ports used. but that is beyond the scope of this book. to use the /etc/exports file to limit access to certain trusted servers or networks.qxd 2/25/05 10:19 AM Page 476 476 Remote Disk Access With NFS Chapter 29 Following these simple but important guidelines will help to make NFS operate more predictably. It can become complicated especially if the users and user groups on the NFS client that are expected to access data on the NFS server don’t exist on the NFS server. File Locking NFS allows multiple clients to mount the same directory.32Harrison_ch29. then they will have access to all the files of a user named nfsuser on the NFS server. although more recent versions are said to have rectified the problem. You may also want to use a firewall to protect access to the NFS server. You may also want to eliminate any wireless networks between your NFS server and client. It is best. Limiting root Access NFS doesn’t allow a root user on a NFS client to have root privileges on the NFS server. but the data is frequently transferred on a randomly chosen TCP port negotiated between them. and it is not wise to mount an NFS share across the Internet as access could be either slow. This is normally an undesirable condition. Nesting Exports NFS doesn’t allow you to export directories that are subdirectories of directories that have already been exported unless they are on different partitions.

CONCLUSION As you have seen NFS can be a very powerful tool in providing clients with access to large amounts of data. Many of the new network-attached storage products currently available on the market rely on NFS—a testament to its popularity. make the key users and groups on both systems match and make sure the permissions on the NFS client mount point and the exported directories of the NFS server are in keeping with your operational objectives.qxd 2/25/05 10:19 AM Page 477 Conclusion 477 For simplicity. .32Harrison_ch29. such as a database stored on a centralized server. increasing stability. and improving security.

qxd 2/25/05 10:19 AM Page 478 .32Harrison_ch29.

but I suggest that you restrict its use to highly secure networks or to networks where access to non-NIS networks is highly restricted.33Harrison_ch30. instead of every system on the network. An advantage of NIS is that users need to change their passwords on the NIS server only. This makes NIS popular in computer training labs. 479 . distributed software development projects or any other situation where groups of people have to share many different computers. The user account is created only on the NIS server. NIS clients download the necessary username and password data from the NIS server to verify each user login.qxd 2/25/05 10:20 AM Page 479 C H A P T E R 30 Centralized Logins Using NIS IN THIS CHAPTER ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ ☞ Scenario Configuring the NFS Server Configuring the NFS Client Configuring the NIS Server Adding New NIS Users Configuring the NIS Client NIS Slave Servers Changing Your NIS Passwords Considerations for a Non-NFS Environment NIS Troubleshooting Conclusion Network Information Services (NIS) enables you to create user accounts that can be shared across all systems on your network. The disadvantages are that NIS doesn’t encrypt the username and password information sent to the clients with each login and that all users have access to the encrypted passwords stored on the NIS server. A detailed analysis of NIS security is beyond the scope of this book.

i386. Edit the /etc/exports file to allow NFS mounts of the with read/write access: /home *(rw. and ypbind and yp-tools are on the client.rpm. ☞ Users logging into the PCs will be assigned home directories on Bigboy and not on the PCs themselves. Downloading and installing RPMs isn’t hard. Configure Smallfry as an NIS client. ☞ The lab instructor will practice with a Linux PC named Smallfry before implementing NIS on all the remaining PCs. Configure Smallfry as an NFS client that can access Bigboy’s /home directory.qxd 2/25/05 10:20 AM Page 480 480 Centralized Logins Using NIS Chapter 30 SCENARIO To understand the benefits of NFS. 2. remember that the filename usually starts with the software package name followed by a version number. ☞ Each user’s home directory will be automatically mounted with each user login on the PCs using NFS.sync) /home directory . Configure Bigboy as an NIS server. has a large amount of disk space and will be used as both the NIS server and NFS-based file server for the Linux PCs in the lab.33Harrison_ch30. Test a remote login from Bigboy to Smallfry using the username and password of the account nisuser.8-3. Bigboy. “Installing RPM Software. 6. Create a user account (nisuser) on Bigboy that doesn’t exist on Smallfry. A school wants to set up a small computer lab for its students: ☞ The main Linux server. Configure Bigboy as an NFS server to make its /home directory available to the Linux workstations. ☞ The suite of NIS RPMs have been installed on the server and client: ypserve and yp-tools are on the server. Convert the account to a NIS user account. consider an example. You have the scenario and the plan. 3. as discussed in Chapter 6. 5. The lab instructor did some research and created an implementation plan: 1. as in yp-tools-2. 4. it’s time to get to work.” When searching for the RPMs. CONFIGURING THE NFS SERVER Here are the steps to configure the school’s NFS server: 1.

These steps archive the /home directory. these prerequisites aren’t necessary. we have to configure its clients. Make sure the required nfs. [root@bigboy tmp]# chkconfig nfslock on [root@bigboy tmp]# chkconfig nfs on [root@bigboy tmp]# chkconfig portmap on [root@bigboy tmp]# service portmap start Starting portmapper: [ OK ] [root@bigboy tmp]# service nfslock start Starting NFS statd: [ OK ] [root@bigboy tmp]# service nfs start Starting NFS services: [ OK ] Starting NFS quotas: [ OK ] Starting NFS daemon: [ OK ] Starting NFS mountd: [ OK ] [root@bigboy tmp]# After configuring the NIS server. restore it to the NFS server. In a production environment in which the /home directory would be actively used. nfslock. As this is a lab environment. and portmap daemons are both running and configured to start after the next reboot. CONFIGURING THE NFS CLIENT You also need to configure the NFS clients to mount their /home directories on the NFS server. This will be covered next. Make sure the required netfs.33Harrison_ch30. back up the data. 1. and then follow the steps below.qxd 2/25/05 10:20 AM Page 481 Configuring the NFS Client 481 2. Let NFS read the /etc/exports file for the new entry. and make available to the network with the exportfs command: [root@bigboy tmp]# exportfs -a [root@bigboy tmp]# /home 3. and portmap daemons are running and configured to start after the next reboot: [root@smallfry tmp]# chkconfig nfslock on [root@smallfry tmp]# chkconfig netfs on [root@smallfry tmp]# chkconfig portmap on [root@smallfry tmp]# service portmap start Starting portmapper: [ OK ] [root@smallfry tmp]# service netfs start Mounting other filesystems: [ OK ] [root@smallfry tmp]# service nfslock start Starting NFS statd: [ OK ] [root@smallfry tmp]# . nfslock. you’d have to force the users to log off.

.1. Edit file /etc/auto.. You’ll be able to test this later after NIS is configured.home to do the NFS mount whenever the /home directory is accessed. .master file to refer to file /etc/auto. Unmount it once everything looks correct: [root@smallfry tmp]# [root@smallfry tmp]# ftpinstall nisuser [root@smallfry tmp]# [root@smallfry tmp]# mount 192. ....100:/home:& 6..wsize=8192. Start configuring autofs automounting.master /home /etc/auto. .nosuid. [root@smallfry tmp]# 3.save .168..home for mounting information whenever the /home directory is accessed.soft.home --timeout 600 5. This is because by default NFS activates the root squash feature.168. autofs unmounts the directory: #/etc/auto.qxd 2/25/05 10:20 AM Page 482 482 Centralized Logins Using NIS Chapter 30 2. Keep a copy of the old /home directory.33Harrison_ch30. [root@smallfry tmp]# mv /home /home. you won’t be able to see the contents of the /home directory on Bigboy as user root. you can add a \ character at the end to continue on the next line: #/etc/auto. which disables this user from having privileged access to directories on remote NFS servers. and create a new directory /home on which you’ll mount the NFS server’s directory.tcp \ 192.save [root@smallfry tmp]# mkdir /home [root@smallfry tmp]# ll / . Make sure you can mount Bigboy’s /home directory on the new /home directory you just created. After five minutes.1. Edit your /etc/auto. Start autofs and make chkconfig command: sure it starts after the next reboot with the [root@smallfry tmp]# chkconfig autofs on [root@smallfry tmp]# service autofs restart Stopping automount:[ OK ] Starting automount:[ OK ] [root@smallfry tmp]# After doing this.rsize=8192.home * -fstype=nfs..intr.100:/home /home/ ls /home quotauser smallfry www umount /home 4. drwxr-xr-x 1 root root 11 Nov 16 20:22 home drwxr-xr-x 2 root root 4096 Jan 24 2003 home. If the line is too long to view on your screen.

0. but requires many steps that you may overlook. The developers had to change the name after a copyright infringement lawsuit. NIS was called Yellow Pages.33Harrison_ch30. CONFIGURING THE NIS SERVER NFS only covers file sharing over the network.conf . Note In the early days. “Remote Disk Access With NFS. when in reality they are automatically mounted and accessed over your network.qxd 2/25/05 10:20 AM Page 483 Configuring the NIS Server 483 Note This automounter feature doesn’t appear to function correctly in my preliminary testing of Fedora Core 3.conf to list the domain’s NIS server as being the server itself or localhost: # /etc/yp.conf File NIS servers also have to be NIS clients themselves. For the school. Don’t worry.1 . The configuration of the NIS server is not difficult.ypbind configuration file ypserver 127. See Chapter 29. Edit Your /etc/sysconfig/network File You need to add the NIS domain you wish to use in the /etc/sysconfig/network file. yet many of the key programs associated with NIS have kept their original names beginning with yp.0. This scheme will make the users feel their home directories are local. we’ll review each one in detail. All newly added Linux users will now be assigned a home directory under the new remote /home directory.” for details. so you’ll have to edit the NIS client configuration file /etc/yp. You now have to configure NIS login authentication for the lab students before the job is done. call the domain NIS-SCHOOL-NETWORK: #/etc/sysconfig/network NISDOMAIN=”NIS-SCHOOL-NETWORK” Edit Your /etc/yp.

[root@bigboy tmp]# service portmap start Starting portmapper: [ OK ] [root@bigboy tmp]# service yppasswdd start Starting YP passwd service: [ OK ] [root@bigboy tmp]# service ypserv start Setting NIS domain name NIS-SCHOOL-NETWORK: Starting YP server services: [ OK ] [root@bigboy tmp]# [root@bigboy [root@bigboy [root@bigboy [root@bigboy tmp]# chkconfig portmap on tmp]# chkconfig yppasswdd on tmp]# chkconfig ypserv on tmp]# [ OK ] Table 30.1 lists a summary of the daemon’s functions. Table 30.33Harrison_ch30. .d directory and use the chkconfig command to ensure they start after the next reboot. You’ll start these daemons after initialization is completed.qxd 2/25/05 10:20 AM Page 484 484 Centralized Logins Using NIS Chapter 30 Start the Key NIS Server-Related Daemons Start the necessary NIS daemons in the /etc/init.1 Required NIS Server Daemons Purpose The foundation RPC daemon upon which NIS runs Lets users change their passwords on the NIS server from NIS clients Main NIS server daemon Main NIS client daemon Used to speed up the transfer of very large NIS maps Daemon Name portmap yppasswdd ypserv ypbind ypxfrd Make sure they are all running before continuing to the next step. You can use the rpcinfo command to do this: [root@bigboy tmp]# rpcinfo -p program vers proto port 100000 2 tcp 111 100000 2 udp 111 100009 1 udp 681 100004 2 udp 698 100004 1 udp 698 100004 2 tcp 701 100004 1 tcp 701 [root@bigboy tmp]# localhost portmapper portmapper yppasswdd ypserv ypserv ypserv ypserv The ypbind and ypxfrd daemons won’t start properly until after you initialize the NIS domain.

.bynumber.aliases... Please continue to add the names for the other hosts. Updating services.....qxd 2/25/05 10:20 AM Page 485 Configuring the NIS Server 485 Initialize Your NIS Domain Now that you have decided on the name of the NIS domain. which in this case is Bigboy.byname.bynumber. Updating protocols.. Running /var/yp/Makefile... Updating hosts. .byname. Updating netid. one per line. Updating group... Updating hosts..byuid... Updating rpc.. gmake[1]: Entering directory `/var/yp/NIS-SCHOOL-NETWORK’ Updating passwd.. bigboy is in the list of NIS server hosts...bygid.byname. When you are done with the list. Updating services.bygid. such as: failed to send ‘clear’ to local ypserv: RPC: Port mapper failureUpdating group.byname. With this procedure. gmake[1]: Leaving directory `/var/yp/NIS-SCHOOL-NETWORK’ bigboy has been set up as a NIS master server.. Updating group. Updating mail. you’ll have to use the ypinit command to create the associated authentication files for the domain. You will be prompted for the name of the NIS server..byname..byservicename... Now you can run ypinit -s bigboy on all slave server.. Updating passwd.. Building /var/yp/NIS-SCHOOL-NETWORK/ypservers..byname. we have to construct a list of the hosts which will run NIS servers...33Harrison_ch30. type a <control D>... all nonprivileged accounts are automatically accessible via NIS: [root@bigboy tmp]# /usr/lib/yp/ypinit -m At this point..byname... Updating protocols. next host to add: bigboy next host to add: The current list of NIS servers looks like this: bigboy Is this correct? [y/n: y] y We need a few minutes to build the databases. Updating rpc... [root@bigboy tmp]# Note Make sure portmap is running before trying this step or you’ll get errors.byaddr.

therefore. are listed using the rpcinfo command when they are running correctly. Start the ypbind and ypxfrd Daemons You can now start the ypbind and the ypxfrd daemons because the NIS domain files have been created: [root@bigboy tmp]# service ypbind start Binding to the NIS domain: [ OK ] Listening for an NIS domain server.33Harrison_ch30. yppasswd. and ypserv before you’ll be able to do this again successfully.qxd 2/25/05 10:20 AM Page 486 486 Centralized Logins Using NIS Chapter 30 You will have to delete the /var/yp/NIS-SCHOOL-NETWORK directory and restart portmap. [root@bigboy tmp]# rpcinfo -p localhost program vers proto port 100000 2 tcp 111 portmapper 100000 2 udp 111 portmapper 100003 2 udp 2049 nfs 100003 3 udp 2049 nfs 100021 1 udp 1024 nlockmgr 100021 3 udp 1024 nlockmgr 100021 4 udp 1024 nlockmgr 100004 2 udp 784 ypserv 100004 1 udp 784 ypserv 100004 2 tcp 787 ypserv 100004 1 tcp 787 ypserv 100009 1 udp 798 yppasswdd 600100069 1 udp 850 fypxfrd 600100069 1 tcp 852 fypxfrd 100007 2 udp 924 ypbind 100007 1 udp 924 ypbind 100007 2 tcp 927 ypbind 100007 1 tcp 927 ypbind [root@bigboy tmp]# . [root@bigboy tmp]# service ypxfrd start Starting YP map server: [ OK ] [root@bigboy tmp]# chkconfig ypbind on [root@bigboy tmp]# chkconfig ypxfrd on Make Sure the Daemons Are Running All the NIS daemons use RPC port mapping and.

. nonprivileged accounts become automatically accessible via NIS.byname.byuid. On a NIS client. gmake[1]: Leaving directory `/var/yp/NIS-SCHOOL-NETWORK’ [root@bigboy yp]# You can check to see if the user’s authentication information has been updated by using the ypmatch command. which has similar syntax..qxd 2/25/05 10:20 AM Page 487 Adding New NIS Users 487 ADDING NEW NIS USERS New NIS users can be created by logging into the NIS server and creating the n